Age-dependent modulation of hippocampal excitability by KCNQ-channels

Epilepsy Research 53 (2003) 81–94

Age-dependent modulation of hippocampal excitability by KCNQ-channels Motohiro Okada a,∗ , Gang Zhu a , Shinichi Hirose b , Ken-Ichi Ito c , Takuya Murakami a , Makoto Wakui d , Sunao Kaneko a a

Department of Neuropsychiatry, Hirosaki University, Hirosaki 036-8562, Japan b Department of Paediatrics, Fukuoka University, Fukuoka 814-0180, Japan c Department of Physiology, Yamagata University, Yamagata 990-9585, Japan d Department of Physiology, Hirosaki University, Hirosaki 036-8562, Japan

Received 17 March 2002; received in revised form 10 November 2002; accepted 17 November 2002

Abstract Recently, mutations of KCNQ2 or KCNQ3, members of the KCNQ-related K+ -channel (KCNQ-channel) family, were identified as cause of benign familial neonatal convulsions (BFNC). However, the exact pathogenic mechanisms of age-dependent development and spontaneous remission of BFNC remain to be elucidated. To clarify the age-dependent etiology of BFNC, we determined age-dependent functional switching of KCNQ-channels, GABAergic- and glutamatergic-transmission in rat hippocampus. The effects of inhibitors of KCNQ-channel, GABA- and glutamate-receptors on propagation of neuronal-excitability and neurotransmitter release were determined by 64-channel multielectrode-dish (MED64), whole-cell recording, in vitro release technique and in vivo microdialysis biosensor, using rat hippocampus from day of birth (P0) to postnatal-day 56 (P56). Inhibition of KCNQ-channels enhanced depolarization-induced glutamate and GABA releases during P0–P7, but not during P14–P28. Inhibition of KCNQ-channels magnified neuronal-excitability propagation from P0 to P14: maximal at P3, but this effect disappeared by P28. GABAA -receptor inhibition surprisingly reduced neuronal-excitability propagation during P0–P3, but not at P7. AMPA/glutamate-receptors inhibition reduced propagation of neuronal-excitability throughout the study period. KCNQ-channels inhibition shortened spike-frequency adaptation, but this stimulation was more predominant during P < 7 than P > 14. During the first week of life, KCNQ-channels performed as a predominant inhibitory system, whereas after this period GABAergictransmission switched from excitatory to inhibitory function. Contrary, glutamatergic-transmission has acquired as excitatory function from P0. These findings suggest that the pathogenic mechanisms of age-dependent development and spontaneous remission of BFNC are, at least partially, associated with the interaction between age-dependent reduction of inhibitory KCNQ-channel activity and age-dependent functional switching of the GABAergic-system from excitatory to inhibitory action in neonatal CNS. © 2002 Elsevier Science B.V. All rights reserved. Keywords: BFNC; Epilepsy; GABA; Glutamate; KCNQ

1. Introduction ∗

Corresponding author. Tel.: +81-172-39-5066; fax: +81-172-39-5067. E-mail address: [email protected] (M. Okada).

Epilepsy is a common neurological disorder afflicting 1–2% of the general population worldwide (Hauser, 1997). Hereditary factors are closely involved

0920-1211/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 0 - 1 2 1 1 ( 0 2 ) 0 0 2 4 9 - 8

82

M. Okada et al. / Epilepsy Research 53 (2003) 81–94

in the etiology of epilepsy (Hirose et al., 2000a) and genetic abnormalities have been identified in a few familial epilepsy syndromes including benign familial neonatal convulsions (BFNC) (Hirose et al., 2000a,b). BFNC is characterized by clusters of generalized seizures exclusively afflicting neonates, with spontaneous remission (ILAE, 1989; Aso and Watanabe, 1992; Bye, 1994; Plouin, 1997). Furthermore, it is well established that seizures of BFNC neonates often include partial seizures (Aso and Watanabe, 1992; Bye, 1994; Plouin, 1997). Several mutations of KCNQ2 and KCNQ3, members of the KCNQ-related K+ -channel (KCNQ-channel) family, have been recently identified to be associated with BFNC (Biervert et al., 1998; Charlier et al., 1998; Singh et al., 1998; Hirose et al., 2000b). Molecular biological experiments demonstrated that KCNQ2, KCNQ3 and KCNQ5 are widely co-expressed almost exclusively in the central nervous system (CNS) including the hippocampus (Biervert et al., 1998; Singh et al., 1998; Schroeder et al., 1998; Cooper et al., 2000; Smith et al., 2001). KCNQ2 and KCNQ5 are thought to fully function, when they are assembled as a heterotetramer with KCNQ3, because KCNQ2/KCNQ3 and KCNQ5/KCNQ3 heterometric channels generate 15- and 5-fold larger current than the corresponding homometric channels, respectively (Biervert et al., 1998; Schroeder et al., 1998, 2000; Tinel et al., 1998; Wang et al., 1998; Cooper et al., 2000; Schwake et al., 2000; Lerche et al., 2000). The major role of the increase in current of heterometric KCNQ-channel is thought to be an increase in surface expression of this channel, since co-expression of KCNQ2 and KCNQ3 led to a large increase in the surface expression of both KCNQ2 and KCNQ3 (Schwake et al., 2000). In addition, recent immunohistological studies indicated that KCNQ2/KCNQ3 heteromeric channel is located on proximal dendrites and soma in human hippocampal pyramidal neuron (Cooper et al., 2000). Taken together with these evidences, the pharmacological and electrophysiological profiles (voltage-dependence and kinetics) of these heterotetramer KCNQ-channels [KCNQ2/KCNQ3 (Biervert et al., 1998; Schroeder et al., 1998; Tinel et al., 1998; Wang et al., 1998; Cooper et al., 2000; Schwake et al., 2000; Smith et al., 2001) and KCNQ3/KCNQ5 (Lerche et al., 2000; Schroeder et al., 2000)] suggest that these channels contribute to the formation of native M-current, which

is an important inhibitory regulator of sub-threshold neuronal excitability in CNS (Zhu et al., 2000). Abnormalities of either KCNQ2 or KCNQ3 identified in BFNC are associated with loss of function of KCNQ-channels (Biervert et al., 1998; Charlier et al., 1998; Singh et al., 1998; Hirose et al., 2000b; Schroeder et al., 1998). Several studies have provided support for the “imbalance hypothesis”, that epileptic seizures are preceded by a relative imbalance between excitatory (i.e. glutamatergic system) and inhibitory (GABAergic system) neurotransmission (Hirose et al., 2000a). Such imbalance consequently precipitates and propagates abnormal neuronal hyperexcitability in the CNS, i.e., epilepsy. Recently, mutations in GABRG2 (encoding GABAA receptor ␥2 subunit) or GABRA1 (encoding GABAA receptor ␣1 subunit) were identified as a cause of generalized epilepsy with febrile seizures (GEFS+) (Baulac et al., 2001), febrile seizures (FS) (Wallace et al., 2001), childhood absence epilepsy (CEA) (Wallace et al., 2001) and juvenile myoclonic epilepsy (JME) (Cossette et al., 2002). Thus, deficient function of mutant KCNQ-channels seems to cause convulsion in BFNC, consistent with the “imbalance hypothesis” (Hirose et al., 2000a). However, the pathogenic mechanisms of the age-dependent development and remission of BFNC during the neonatal period remain to be explained in relation to dysfunction of KCNQ-channels (Hirose et al., 2000a). In addition, although long term prognosis of BFNC is considered benign, individuals with BFNC have a higher risk for subsequent epilepsies (11%) compared with the general population (Plouin, 1997). To investigate the possible relationship between deficient KCNQ-channels and age-dependent etiology of BFNC, including development, spontaneous remission and propensity for subsequent epilepsies in BFNC, the present study determined the age-dependent functional switching of KCNQ-channels, GABAergic and glutamatergic transmission in immature rat hippocampus.

2. Materials and methods All of the experiments described in this report were performed in accordance with the specifications of the Ethical Committee of Hirosaki University and met the guidelines of the responsible governmental agency.

M. Okada et al. / Epilepsy Research 53 (2003) 81–94

Wistar rats (Clea, Tokyo, Japan) were housed under conditions of constant temperature at 22 ± 2 ◦ C with a 12-h light:12-h dark cycle. 2.1. Preparation of MED64 system The procedures for preparation of rat hippocampal slices (350 ␮m) from birth (P0) to postnatal days 28 (P28) and 64-channel multielectrode dish (MED64) system (Alpha MED Sciences, Tokyo, Japan), a novel two-dimensional neuronal electroactivity monitoring technique, were based on the methods described by Edwards et al. (1989) and Oka et al. (1999), respectively. We used two types of MED64 probes for determination of hippocampal evoked field potentials, including the following measurements: fiber volley (FV), which is an indicator of presynaptic intraneuronal response; field excitatory post synaptic potentials (fEPSP), an indicator of postsynaptic AMPA/glutamate receptor response; population spikes (PS), an indicator of depolarization of postsynaptic neurons. MED-P5155 (Alpha MED Sciences), an 8 × 8 array with 150 ␮m interpolar distance, was used for hippocampal slices obtained from rats at P0–P7. MED-P5305 (Alpha MED Sciences), with 300 ␮m interpolar distance, was used for those at P14–P28, since MED-P5155 could not cover the hippocampal slices obtained from rats at P14–P28 (Zhu et al., 2000). During electrophysiological recording, the hippocampal slice positioned on the MED64 probe (Fig. 1A) was superfused with artificial cerebrospinal fluid (ACSF) composed of (in mM): 150 Na+ , 2.5 K+ , 2.0 Ca2+ , 1.0 Mg2+ , 132.25 Cl− and buffered with 1.25 mM phosphate buffer and 25 mM carbonate buffer to adjust pH to 7.35 at 35 ◦ C, and containing 10 ␮M Dup996 (KCNQ-channel inhibitor), 40 ␮M DNQX (AMPA/glutamate receptor inhibitor) or 10 ␮M bicuculline (GABAA receptor inhibitor), at a flow rate of 2 ml/min. A pair of single planar microelectrodes with bipolar constant current pulses (10–100 ␮A, 0.1 ms) was used for stimulation of the hippocampal slice. The stimulation site was selected at the Schaffer collateral pathway in the CA1 region, and both propagation area and amplitude of FV, fEPSP and PS were recorded with all 64-planar microelectrodes, which covered the hippocampal slice (Edwards et al., 1989; Oka et al., 1999; Zhu et al., 2000).

83

2.2. In vitro neurotransmitter release From P0 to P28, hippocampal mini-slices (350 ␮m× 350 ␮m) prepared on a chopper were placed in modified ACSF (M-ACSF) composed of (in mM): 145 Na+ ; 2.7 K+ ; 1.2 Ca2+ ; 1.0 Mg2+ ; 145.1 Cl− and buffered with 1.5 mM phosphate buffer and 10 mM HEPES to adjust pH to 7.35. After incubation, the slices were extensively washed with M-ACSF and placed in baskets (150 ␮m nylon mesh bases) to transfer them between the washing and release buffers. Transmitter release studies were carried out in quadruplicate at 37 ◦ C using the discontinuous method. To measure the releases of glutamate and GABA, 20 mg of wet weight hippocampal tissue (mg wwht) were incubated in 0.5 ml of M-ACSF. To obtain a stable basal release, slices were incubated for 11 min in M-ACSF, transferred to a second tube containing M-ACSF and further incubated for 4 min. Following the latter incubation, the slices were transferred to either M-ACSF or M-ACSF containing 50 mM K+ (K-ACSF) in which Na+ was substituted on an equimolar basis to achieve a high concentration of K+ (K+ -evoked release), and incubation was carried out for 4 min. To determine the release level of each neurotransmitter, after the incubation, M-ACSF or K-ACSF was injected into high performance liquid chromatography (HPLC) systems (Zhu et al., 2000; Kawata et al., 2001; Okada et al., 2001). The extracellular amino acids levels were determined by HPLC with fluorescence detection. The analytical column (100 mm × 1.5 mm i.d.) was packed with mightysil RP-18 (particle size, 3 ␮m), which was purchased from Kanto Chemicals (Tokyo), by Masis INC (Hirosaki) (Murakami et al., 2001; Okada et al., 2001). The excitation and emission fluorescence wavelengths were 340 and 445 nm, respectively. The mobile phase was 0.1 M phosphate buffer (pH 6.0) containing 20% methanol and a flow rate of 300 ␮l/min (Zhu et al., 2000). 2.3. Whole-cell recording Hippocampal slices (400 ␮m thick) were prepared from male Wistar rats (Clea, Tokyo). The standard solution (pH 7.4) contained (in mM): NaCl, 124; KCl, 3.0; CaCl2 , 2.5; MgSO4 , 2.0; NaHCO3 , 22.0; NaH2 PO4 , 1.25; and glucose, 10.0; and was bubbled constantly with a gas mixture of 5% CO2 –95% O2 .

84

M. Okada et al. / Epilepsy Research 53 (2003) 81–94

Fig. 1. Effects of blockers of KCNQ-channel, GABA and AMPA/glutamate receptors on propagation of FV and fEPSP in hippocampal CA1 regions during P0–P28. A typical phase-contrast photomicrograph of a rat hippocampal slice at P14 positioned on MED64 probe (MED-P5305: scale bar = 300 ␮m) is shown in (A). The stimulation site was selected at the Schaffer collateral pathway in the CA1 region (red arrow), and FV, fEPSP and PS were recorded with all 64-planar-microelectrodes, which covered the whole hippocampal slice. The amplitude and propagation of each parameter recorded with the electrodes of the MED64 system are expressed two-dimensionally; amplitude is depicted in colors according to the color chart shown in the legend and area represents the degree of propagation. Response of Schaffer collateral evoked FV (B) and fEPSP (C) to 10 ␮M Dup996, 10 ␮M bicuculline and 40 ␮M DNQX in hippocampal slices at the indicated postnatal days. The hippocampal slices, during P0–P7 and P14–P28 were placed on MED-P5155 and MED-P5305, respectively, which were arranged in an 8 × 8 array with an interpolar distance of 150 and 300 ␮m, respectively. The effects of Dup996, bicuculline and DNQX on propagation and amplitude of FV, fEPSP and PS were analyzed by one-way ANOVA with Tukey’s multiple comparison.

Slices were incubated in this medium for at least 60 min prior to the first recording session. Patch pipettes were prepared by heating and pulling borosilicate glass capillaries (1.5/1.12 mm o.d./i.d.; WPI, Sarasota, FL) using a double-stage puller (Narishige, Tokyo, Japan). The tip diameter was 1–2 ␮m and the tip resistance was 5–7 M. The pipette was filled with a solution containing (in mM): potassium gluconate, 142; HEPES, 10; EGTA, 0.5; Mg-ATP, 2; MgCl2 ·6H2 O, 1; NaCl, 10; and GTP-2Na, 0.2. The

pH was adjusted to 7.3 with KOH and the osmolarity was 285–300 mOsmol/l. The pipette series resistance was 20–35 M. and was not compensated so as to maintain the signal-to-noise ratio as high as possible. The series resistance was monitored throughout the experiment and if it increased by >15%, the experiment was excluded from further analysis. The following criteria were used to determine the suitability of a CA1 pyramidal neuron for recording: a negative resting membrane potential (RMP) <−55 mV,

M. Okada et al. / Epilepsy Research 53 (2003) 81–94

input resistance >20 M (as measured at RMP by a 500 pA hyperpolarizing pulse). The RMP and number of spikes were determined using Axopatch 200A (Axon Instruments) and stored on computer using custom-written software (Ito et al., 1999). 2.4. In vivo microdialysis biosensor In vivo microdialysis biosensor consisted of a dialysis-electrode (250 ␮m diameter, 2 mm exposed membrane; Sycopel, London, UK) coated with O-phenylenediamine, a dialysis probe (220 ␮m diameter, 2 mm exposed membrane; EICOM, Kyoto, Japan) and teflon-coated twin stainless steel recording electrodes (80 ␮m diameter and 100 ␮m space) (Okada et al., 1998). The dialysis electrode was set at a potential of 450 mV and filled with modified Ringer’s solution (MRS-C) composed of (in mM): 145 Na+ , 2.7 K+ , 1.2 Ca2+ , 1.0 Mg2+ , 125.9 Cl− , 1.2 PO4 3− and 25 CO3 2− (pH 7.40), containing glutamate oxidase (100 U/ml, Yamasa, Chiba, Japan). Male Wistar rats weighing 250–300 g, were anaesthetized with 1.0 g/kg urethane injected subcutaneously, placed in a stereotaxic frame and body temperature was kept at 37 ◦ C using a heating pad. The in vivo microdialysis biosensor was implanted into the hippocampus (A = −5.8 mm, L = 4.8 mm, V = −4.0 mm, relative to the bregma) and perfusion experiments commenced at least 6 h after the hippocampal extracellular glutamate level had reached a plateau (Okada et al., 1998). To study the effects of increased extracellular K+ (K+ -evoked stimulation) on hippocampal extracellular levels of glutamate, MRS-C containing 50 or 100 mM K+ (HK-MRS) was perfused for 60 min (Okada et al., 1998). The ionic composition was modified and isotonicity was maintained by equimolar decrease of Na+ (Okada et al., 1998). Monitoring of neuronal firing frequency was performed as described previously by our laboratory (Okada et al., 1998). Neuronal firing frequency was recorded by a telemetry system (Unimec Co., Tokyo), with the bandpass filter set at 0.1–3 kHz, and digitized as the discharge rate. 2.5. Statistical analysis Data are expressed as mean ± S.D. The age-dependent responses of FV, fEPSP and PS to Dup996, bicu-

85

culline and DNQX were analyzed by repeated measurements ANOVA or Freedman’s test. The effects of Dup996 on the number of spikes and RMP were analyzed by Wilcoxon test and paired t-test, respectively, and that on the inter-spike intervals were repeated measurements ANOVA. The effects of Dup996 on the responses of basal and K+ -evoked release were analyzed by one-way ANOVA. All ANOVA tests were followed by Tukey’s multiple comparison test. A P value less than 0.05 denotes the presence of a statistically significant difference.

3. Results 3.1. Propagation of neuronal excitability To investigate the possible relationship between KCNQ-deficient channels and the age-dependent etiology of BFNC, we examined the effects of KCNQ-channel, GABAergic and glutamatergic transmission system on propagation of neuronal excitability using MED64 system (Alpha MED Sciences), a novel two-dimensional neuronal electroactivity monitoring technique (Oka et al., 1999; Shimono et al., 2000; Zhu et al., 2000). To clarify the role of these neurotransmissions, each type of channel or receptor was inhibited with a selective chemical blocker. Especially, Dup996 was used at a concentration of 10 ␮M to induce approximately 50% reduction in hippocampal KCNQ related M-current (Aiken et al., 1995), which was composed of KCNQ2/KCNQ3 (Biervert et al., 1998; Schroeder et al., 1998; Wang et al., 1998) and KCNQ3/KCNQ5 (Lerche et al., 2000; Schroeder et al., 2000) heterometric channels, without affecting other types of K+ -current including Erg-related M-like K+ -current (Lamas et al., 1997; Schnee and Brown, 1998; Meves et al., 1999; Selyanko et al., 1998) mimicking currents obtained from deficient KCNQ-channel in BFNC (Biervert et al., 1998; Schroeder et al., 1998; Tinel et al., 1998; Wang et al., 1998; Cooper et al., 2000; Schwake et al., 2000). The effects of such inhibition were then evaluated in rat hippocampal CA1 regions from P0 to P28. The parameters analyzed in the study were the amplitude and propagation of evoked field potentials (FP) including FV, fEPSP and PS. When the signal-to-noise (S/N) ratio was more than 3, we recorded Schaffer

86

M. Okada et al. / Epilepsy Research 53 (2003) 81–94

collateral-evoked FV, fEPSP and PS as the detectable responses. The parameter of “propagation” means the total number of electrodes, which could monitor the detectable responses. The parameter of “amplitude” means the total amount of amplitude of each FV, fEPSP and PS recorded by all electrodes, which could monitor the detectable responses. During P0–P28, Dup996 did not affect the amplitude of FV, fEPSP or PS (Figs. 1 and 2), whereas both propagation of FV (P < 0.01, N = 6) and fEPSP (P < 0.01, N = 6) were magnified with Dup996 dur-

ing P0–P14 (Figs. 1 and 2). The stimulatory effect of Dup996 on FV and fEPSP propagation was maximal at P3 (Figs. 1 and 2) but disappeared at P28 (Figs. 1 and 2). Surprisingly, inhibition of GABAA receptor by 10 ␮M bicuculline, a selective GABAA receptor inhibitor, reduced the amplitude of FV (P < 0.01, N = 6), fEPSP (P < 0.01, N = 6) and PS (P < 0.01, N = 6), and reduced the propagation of FV (P < 0.01, N = 6) and fEPSP (P < 0.01, N = 6) during P0–P3, but not at P7 (Figs. 1 and 2). On the other hand, during

Fig. 2. Effects of blockers of KCNQ-channel, GABA and AMPA/glutamate receptors on propagation and amplitude of FV, fEPSP and PS in hippocampal CA1 regions during P0–P28. Age-dependent responses of Schaffer collateral-evoked FV, fEPSP and PS to 10 ␮M Dup996 (䊊), 10 ␮M bicuculline (䊉) and 40 ␮M DNQX (䊐) in hippocampal slices. The ordinates indicate the percent control (perfusion without each agent) value of amplitude of FV (A), fEPSP (C) and PS (E) or propagation of FV (B) and fEPSP (D). The abscissas indicate the postnatal days. The age-dependent responses of Schaffer collateral-evoked FV, fEPSP and PS to Dup996, bicuculline and DNQX were analyzed by repeated measurements ANOVA or Freedman test with Tukey’s multiple comparison (∗∗ P < 0.01).

M. Okada et al. / Epilepsy Research 53 (2003) 81–94

P14–P28, bicuculline enhanced the amplitude of FV, fEPSP and PS, and enhanced the propagation of FV and fEPSP (Figs. 1 and 2). In the next series of experiments, we blocked AMPA/glutamate receptors by using 40 ␮M DNQX, a selective inhibitor. Such inhibitor reduced the amplitude of fEPSP (P < 0.01, N = 6) and PS (P < 0.01, N = 6), and reduced fEPSP propagation (P < 0.01, N = 6) without affecting the amplitude or propagation of FV throughout the study period (Figs. 1 and 2). 3.2. Neurotransmitter release in vitro The developmental changes in the role of KCNQchannels, and the interaction between GABAergic

87

and glutamatergic transmissions demonstrated above appeared to be associated with functional maturation of these channels or receptors themselves. However, developmental changes in neurotransmitter release should also be considered for GABA and AMPA/glutamate receptors. For this purpose, we measured both basal and 50 mM K+ -evoked release of glutamate and GABA in hippocampal mini-slices in vitro. The amount of basal glutamate release decreased in an age-dependent manner (P < 0.05, N = 6), whereas that of GABA increased (P < 0.01, N = 6) with maturation (Fig. 3), indicating that GABA release at resting stage continues to develop with growth. Dup996 (10 ␮M), which reduced KCNQ-channel activity (Figs. 1 and 2), did not influence basal release

Fig. 3. Effects of KCNQ-channel inhibitor on basal and K+ -evoked fractional releases of glutamate and GABA in rat hippocampal mini slices in vitro. Effects of 10 ␮M Dup996 on basal and K+ -evoked glutamate (A) and GABA (B) releases in hippocampal slices in vitro. Left ordinate indicates 50 mM K+ -evoked release of glutamate and GABA (nmol/g wwht), while the right ordinate indicates the basal release of each neurotransmitter (nmol/g wwht). Open (䊊) and closed (䊉) circles indicate releases under incubation with M-ACSF containing 50 mM K+ with and without 10 ␮M Dup996 for 4 min, respectively. Open (䊐) and closed (䊏) squares indicate releases under incubation with M-ACSF with and without 10 ␮M Dup996, for 4 min, respectively. Responses of basal and K+ -evoked release to Dup996 at the indicated postnatal days were analyzed by one-way ANOVA with Tukey’s multiple comparison (∗ P < 0.05; ∗∗ P < 0.01, compared to mini slices superfused with M-ACSF free of Dup996).

88

M. Okada et al. / Epilepsy Research 53 (2003) 81–94

of glutamate or GABA throughout P0–P28 period (Fig. 3). On the other hand, K+ -evoked release of glutamate (P < 0.01, N = 6) and GABA (P < 0.01, N = 6) increased age-dependently (Fig. 3). Dup996 enhanced such K+ -evoked releases of both glutamate and GABA during P0–P7, but not during P14–P28 (Fig. 3). 3.3. Resting membrane potential and spike frequency adaptation To clarify the stimulatory effects of Dup996 on propagation of neuronal excitability and depolarisationrelated neurotransmitter release, we monitored the resting membrane potential (RMP) and spike frequency adaptation (SFA) in rat CA1 hippocampal neuron, using the slice patch-clamp technique (Ito et al., 1999). Within the first week of life (P < 7), Dup996 (10 ␮M) shortened SFA and increased the number of spikes (136.3 ± 14.3%; N = 6; P < 0.01) (Fig. 4). During P14–P28 (P > 14), Dup996 (10 ␮M) also shortened SFA and increased the number of spikes (115.8 ± 8.5%; N = 6; P < 0.05). These

stimulatory effects of Dup996 on SFA were more predominant during P < 7 than P > 14 (P < 0.01) (Fig. 4). Dup996 (10 ␮M) significantly elevated RMP from −64.4 ± 2.8 to −61.8 ± 3.1 mV (P < 0.05) using paired t-test, but this elevation was trivial (ranged 2–4 mV) (data not shown). The response of RMP to Dup996 between P < 7 and P > 14 was not different (data not shown). 3.4. Spreading depression related neurotransmitter release in mature CNS To study the relationship between deficient KCNQchannels and propensity for subsequent epilepsies in BFNC, the effects of reduced KCNQ related M-current on neurotransmitter release and the frequency of neuronal firing (NF) were investigated in mature rat CNS (P56). For this purpose, glutamate release and NF frequency were monitored (GABA release could not be determined) using an in vivo microdialysis biosensor for real-time monitoring (Okada et al., 1998) either at rest or hyperexcitation (50 and 100 mM K+ -evoked stimulation). To mimic impaired

Fig. 4. Effects of KCNQ channel inhibitor on SFA in CA1 hippocampal neuron. (A and B) Effects of Dup996 on the number of spikes, analyzed by Wilcoxon test (∗ P < 0.05; ∗∗ P < 0.01). Data indicate mean ± S.D. of percent control (treatment without Dup996) of number of spikes during 500 pA hyperpolarizing pulse for 300 ms. (C and D) Effects of 10 ␮M Dup996 on response of a representative CA1 hippocampal neuron to a pulse of depolarisation current (500 pA) at P < 7 (C) and P > 14 (D). Ordinate indicates the number of spikes.

M. Okada et al. / Epilepsy Research 53 (2003) 81–94

89

Fig. 5. Real-time monitoring of effects of KCNQ channels inhibitor on multicomponents of K+ -evoked glutamate release and NF frequency using an in vivo microdialysis biosensor. Effects of modified Ringer’s solution containing 100 mM K+ without (A) or with (B) 10 ␮M Dup996 on hippocampal glutamate releases (lines) and NF frequency (closed columns) examined using the in vivo microdialysis biosensor. Top ordinates indicate extracellular glutamate level (␮M), bottom ordinates indicate NF frequency (Hz/min). Abscissas indicate time in minutes (min). Open bars indicate perfusion with Dup996, and hatched bars indicate 100 mM K+ -evoked stimulation.

KCNQ related M-current in BFNC, KCNQ-channels were blocked by perfusion with 100 ␮M Dup996 (Dup996 concentration in brain tissue was estimated to be 8–11 ␮M, since Dup996 diffusion rate from dialysis probe to brain tissue was 8–11%) (Pieniaszek et al., 1995). During resting stage, Dup996 did not alter basal glutamate release or NF frequency (time, −5 min; Fig. 5). Furthermore, Dup996 again did not alter 50 mM K+ -evoked glutamate release or NF frequency (data not shown). However, the 100 mM K+ -evoked stimulation (60 min) increased glutamate release in a triphasic pattern (Okada et al., 1998). Two phases included an initial transient rise (ITR) and a late gentle rise (LGR), and both were NF frequencydependent (Okada et al., 1998). The third component was a late multiphasic transient rise (LMPTR), a surge of neurotransmitter synchronized with “spreading depression” (Okada et al., 1998) (Fig. 5). The 100 mM K+ -evoked increase in NF frequency and amount of glutamate release during ITR, LGR and LMPTR were enhanced by Dup996 (Fig. 5). Dup996 also increased

the frequency of LMPTR and spreading depression accordingly under the provocative condition (Fig. 5).

4. Discussion Several studies on mutant KCNQ-channels suggest that none of the identified mutations in BFNC exert dominant negative effects, and consequently the reduction of M-current in BFNC patients was predicted to be small (Marrion, 1997; Biervert et al., 1998; Charlier et al., 1998; Singh et al., 1998; Schroeder et al., 1998; Schwake et al., 2000). A KCNQ2 mutant associated with BFNC (1600ins5) that has a truncated cytoplasmic carboxyl terminus did not reach the surface and failed to stimulate KCNQ3 surface expression (Schwake et al., 2000). However, several other BFNC associated missense mutations in KCNQ2 or KCNQ3 did not alter their surface expression (Schwake et al., 2000). Therefore, the pathogenic mechanisms of BFNC are considered to be a dysfunction of the

90

M. Okada et al. / Epilepsy Research 53 (2003) 81–94

inhibitory neurotransmission system induced by mutant KCNQ-channels which reduce seizure threshold and/or enhancement of neuronal excitability (Hirose et al., 2000a). However, one of the most important clinical features of BFNC, which is the age-dependent development and spontaneous remission of BFNC (ILAE, 1989; Aso and Watanabe, 1992; Bye, 1994; Plouin, 1997), cannot be explained by only a dysfunctional mutant KCNQ-channel (Hirose et al., 2000a,b; Zhu et al., 2000). It is well established that native M-current maintains the neuronal stabilization via regulation of neuronal excitability (Marrion, 1997). M-current inhibitors increased stimulus-evoked neurotransmitter release (Aiken et al., 1995; Tam and Zaczek, 1995; Noda et al., 1998; Zhu et al., 2000). Cooper and his colleagues recently suggested that KCNQ2/KCNQ3 channel, which is probably not associated with NMDA/glutamate receptor, regulated neuronal excitability by controlling the ability of excitatory postsynaptic potentials received at dendritic branches and spines to spread centrally and generate action potentials (Cooper et al., 2000). The present study clearly demonstrated both the inhibitory function of KCNQ-channel and its age-dependent functional switching from birth to P28. Within the first week of life, two-dimensional monitoring of neuronal activity demonstrated that the inhibition of KCNQ-channel activity enhanced propagation of FV, an indicator of presynaptic intraneuronal response and that of fEPSP, an indicator of postsynaptic AMPA/glutamate-receptor response without affecting these amplitudes; whereas after this period, its inhibition was not observed. In the patch-clamp experiments, inhibition of KCNQ-channel elevated RMP slightly and altered neuronal firing patterns as evidenced by a shortening of SFA. This reduction of SFA weakened with maturation, but the elevation of RMP was not affected by maturation. Therefore, these electrophysiological data indicate that in immature CNS, KCNQ-channel prevents propagation of neuronal excitability, whereas its inhibition disappears after this period. Furthermore, within the first week of life, inhibition of KCNQ-channel activity increased K+ -evoked neurotransmitter release without affecting basal release. However, after this period, KCNQ-channel affected neither basal nor depolarisation-related releases. These results suggest that in immature CNS, the elevation of RMP induced

by inhibition of KCNQ-channel activity cannot generate the action potentials, resulting in no change in the amount of basal neurotransmitter release. However, KCNQ-channel suppresses the gross count of responsible synapses, which can release glutamate and GABA without affecting the amount of neurotransmitter release per synapse. Furthermore, the present results suggest that within the first week of life, KCNQ-channel maintains neuronal stabilization via both presynaptic and postsynaptic inhibitory regulation during neuronal hyperexcitability, because the inhibition of KCNQ-channel enhanced propagation of FV and fEPSP, increased depolarisation-related neurotransmitter release, and reduced SFA, but did not affect basal neurotransmitter release. Recent studies using northern blot analysis demonstrated that expression of KCNQ2 rapidly increases within the first week of life in rat CNS, followed by that of KCNQ3 (Tinel et al., 1998; Smith et al., 2001). However, during this period when the brain tissue is comprised of 90% glia cells and 10% neuronal cells (Bacci et al., 1999), KCNQ2 is expressed in neuronal cells but not glia cells (Cooper et al., 2000; Smith et al., 2001). Taken together with these evidences, the present results indicate that in immature CNS, KCNQ-channel plays an important role in the inhibitory neurotransmission system despite the low, though increasing, density of KCNQ-channel. To clarify this contradiction between level of mRNA expression and pharmacological functional evidence of KCNQ-channel, the density of KCNQ-channel per neuron should be determined, since increased density of KCNQs mRNA may be dependent on increased number of neurons. Nevertheless, the pathogenesis of BFNC cannot be explained only by the dysfunction of mutant KCNQ-channel but also requires the involvement of another neurotransmission system. Results of a number of studies provide support for the “imbalance hypothesis”, that epileptic seizures are preceded by a relative imbalance between excitatory and inhibitory neurotransmission (Hirose et al., 2000a). Such imbalance consequently precipitates and propagates abnormal neuronal hyperexcitability in the CNS, i.e., epilepsy (Hirose et al., 2000a). Our study showed that glutamatergic transmission has already developed at birth, since both AMPA/glutamate mediated excitatory inputs and depolarisation related glutamate release did not change with maturation.

M. Okada et al. / Epilepsy Research 53 (2003) 81–94

The patch-clamp analysis also demonstrated that NBQX, another selective AMPA/glutamate receptor antagonist, inhibited fEPSP, and the NBQX potency was reduced age-dependently (Seifert et al., 2000). The “imbalance hypothesis” was recently demonstrated in gene analysis of human epileptic patients (Baulac et al., 2001; Wallace et al., 2001). Mutations in GABRG2 (encoding GABAA receptor ␥2 subunit) were identified as a cause of GEFS+ (Baulac et al., 2001), FS (Wallace et al., 2001) and CEA (Wallace et al., 2001). Analysis of these mutated and wild-type genes in Xenopus oocytes confirmed the predicted effects of mutation, a decrease in amplitude of GABA-activated currents (Baulac et al., 2001) and insensitivity to benzodiazepine, a major antiepileptic agent (Wallace et al., 2001). In addition, mutation in GABRA1 (encoding GABAA receptor ␣1 subunit) has been found in a large family with JME (Cossette et al., 2001). Analysis of this mutated and wild-type genes in HEK 293 cells confirmed the predicted effects of mutation, a decrease in amplitude of GABA-activated currents (Cossette et al., 2001). Surprisingly, however, our results indicated that GABA acts as an excitatory neurotransmitter at early stage of development (during the first week of life), and acquires inhibitory function after this period. This observation is supported by a report indicating that GABAA receptor mediated depolarisation was blocked by bicuculline during embryonic and early postnatal period (Owens et al., 1996; Ganguly et al., 2001). The amount of basal GABA release increased age dependently, indicating that at resting stage, basal GABA release continues to develop with growth. The excitatory characteristics of GABAA receptors observed within the first week of life, may not manifest fully due to the immature GABA release response at the resting stage. Contrary to basal release, K+ -evoked GABA release increased age dependently, indicating that depolarisation-induced GABA release continues to develop with maturation. Interestingly, during the first week of life, when GABAA receptor is acting as an excitatory neurotransmission system, KCNQ-channel acts as a predominant inhibitory neurotransmission system. During this period, reduction of KCNQ-channel activity enhances depolarisation-related releases of excitatory neurotransmitters, glutamate and GABA. In immature CNS, under the condition of mutant KCNQ-channel

91

and GABAergic excitatory function, once neuronal hyperexcitability occurs, mutant KCNQ-channel cannot inhibit the propagation of this excitation, and consequently, epileptic seizure is induced by synchronization between increased release of GABA and glutamate from presynaptic terminal and activation of excitatory glutamate and GABAA receptors. Therefore, in immature CNS, this synchronization between the impaired M-current and excitatory GABAergic function, which is expected to be in BFNC, can result in neuronal hyperexcitability. Beyond this period, GABAA receptors switch to an inhibitory function and the age-dependent increase in GABA release at both resting and excitability stages ensures an inhibitory function of mature GABAergic transmission. Thus, the mechanisms for development and remission of BFNC may be associated with both an impaired M-current and a functional switching of GABAergic transmission, i.e., from excitatory to inhibitory function. Hence, the balance between excitatory and inhibitory neurotransmission in such immature CNS should pivot on KCNQ related M-current. A small reduction in KCNQ related M-current resulting from deficient KCNQ-channels in BFNC may lead to convulsions only during neonatal age. This hypothesis is substantiated by the following observation. The electroencephalogram (EEG) showed no apparent differences between wild-type and heterogeneous KCNQ2 knock-out mice; however, heterogeneous KCNQ2 knock-out mice showed hypersensitivity to a proconvulsant, pentylentetrazole (Watanabe et al., 2000). The long term prognosis of BFNC is considered as a benign disorder because most patients with BFNC show age-dependent spontaneous remission. Yet, individuals with BFNC have a higher risk for subsequent epilepsies (11%) compared with the general population (Plouin, 1997). The present results showed that inhibition of KCNQ-channel did not affect 50 mM K+ -evoked glutamate release or NF, but enhanced 100 mM K+ -evoked glutamate release, NF and the frequency of spreading depression, which are neuronal damage-related electrophysiological and neurotransmitter release events (Walker et al., 1995; Obrenovitch and Zilkha, 1996; Okada et al., 1998). Spreading depression, which was originally observed as an electrophysiological phenomenon characterized by a transient suppression of neuronal excitability during consecutive neuronal hyperactivity (Walker et al.,

92

M. Okada et al. / Epilepsy Research 53 (2003) 81–94

1995; Obrenovitch and Zilkha, 1996; Okada et al., 1998) is considered to be associated with neuronal damage in epilepsy (Walker et al., 1995; Okada et al., 1998). Several antiepileptic agents suppress spreading depression and the related neurotransmitter release (Okada et al., 1998). Under neuronal hyperexcitability conditions where aberrant provocative inputs take place, reduced KCNQ related M-current may precipitate seizure activity. Therefore, KCNQ-channels act to stabilize neuronal excitability to provocative inputs, however, the dysfunctional mutant KCNQ-channel associated with BFNC can not only suppress the tonic neuronal hyperexcitability but produces neuronal disturbance, i.e., spreading depression. In conclusion, we have demonstrated that the agedependent development and spontaneous remission of BFNC are produced by cooperation between dysfunctional KCNQ-channel, which plays an important role as a predominant inhibitory neurotransmission system, and developmental functional switching of GABAergic transmission from excitatory to inhibitory in immature CNS.

Acknowledgements This study was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (05454309, 11770532, 13670979 and 13770532), grants from Hirosaki Research Institute for Neurosciences, Pharmacopsychiatry Research Foundation, The Epilepsy Research Foundation, Uehara Memorial Foundation, Heiwa Nakajima Foundation, International Research Fund of Kyushu University School of Medicine Alumni, The Clinical Research Foundation, The Foundation for the Advancement of Clinical Medicine, Japan and The Central Research Institute of Fukuoka University.

References Aiken, S.P., Lampe, B.J., Murphy, P.A., Brown, B.S., 1995. Reduction of spike frequency adaptation and blockade of M-current in rat CA1 pyramidal neurones by linopirdine (DuP 996) a neurotransmitter release enhancer. Br. J. Pharmacol. 115, 1163–1168.

Aso, K., Watanabe, K., 1992. Benign familial neonatal convulsions: generalized epilepsy? Pediatr. Neurol. 8, 226–228. Bacci, A., Verderio, C., Pravettoni, E., Matteoli, M., 1999. The role of glial cells in synaptic function. Phil. Trans. R. Soc. Lond. B. Biol. Sci. 354, 403–409. Baulac, S., Huberfeld, G., Gourfinkel-An, I., Mitropoulou, G., Beranger, A., Prud’homme, J.F., Baulac, M., Brice, A., Bruzzone, R., LeGuern, E., 2001. First genetic evidence of GABAA receptor dysfunction in epilepsy: a mutation in the ␥2 -subunit gene. Nat. Genet. 28, 46–48. Biervert, C., Schroeder, B.C., Kubisch, C., Berkovic, S.F., Propping, P., Jentsch, T.J., Steinlein, O.K., 1998. A potassium channel mutation in neonatal human epilepsy. Science 279, 403–406. Bye, A.M., 1994. Neonate with benign familial neonatal convulsions: recorded generalized and focal seizures. Pediatr. Neurol. 10, 164–165. Cossette, P., Liu, L., Brisebois, K., Dong, H., Lortie, A., Vanasse, M., Saint-Hilaire, J.M., Carmant, L., Verner, A., Lu, W.Y., Tian Wang, Y., Rouleau, G.A., 2002. Mutation of GABRA1 in an autosomal dominant form of juvenile myoclonic epilepsy. Nat. Genet. 31, 184–189. Charlier, C., Singh, N.A., Ryan, S.G., Lewis, T.B., Reus, B.E., Leach, R.J., Leppert, M., 1998. A pore mutation in a novel KQT-like potassium channel gene in an idiopathic epilepsy family. Nat. Genet. 18, 53–55. Commission on Classification and Terminology of the International League Against Epilepsy, 1989. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 30, 389–399. Cooper, E.C., Aldape, K.D., Abosch, A., Barbaro, N.M., Berger, M.S., Peacock, W.S., Jan, Y.N., Jan, L.Y., 2000. Colocalization and coassembly of two human brain M-type potassium channel subunits that are mutated in epilepsy. Proc. Natl. Acad. Sci. U.S.A. 97, 4914–4919. Edwards, F.A., Konnerth, A., Sakmann, B., Takahashi, T., 1989. A thin slice preparation for patch clamp recordings from neurones of the mammalian central nervous system. Pflügers Arch. 414, 600–612. Ganguly, K., Schinder, A.F., Wong, S.T., Poo, M., 2001. GABA itself promotes the developmental switch of neuronal GABAergic responses from excitation to inhibition. Cell 105, 521–532. Hauser, W.A., 1997. Incidence and prevalence. In: Engel Jr., J., Pedley, T.A. (Eds.), Epilepsy: A Comprehensive Textbook. Lippincott-Raven, Philadelphia, pp. 47–57. Hirose, S., Okada, M., Kaneko, S., Mitsudome, A., 2000a. Are some idiopathic epilepsies disorders of ion channels: a working hypothesis. Epilepsy Res. 41, 191–204. Hirose, S., Zenri, F., Akiyoshi, H., Fukuma, G., Iwata, H., Inoue, T., Yonetani, M., Tsutsumi, M., Muranaka, H., Kurokawa, T., Hanai, T., Wada, K., Kaneko, S., Mitsudome, A., 2000b. A novel mutation of KCNQ3 (c.925T → C) in a Japanese family with benign familial neonatal convulsions. Ann. Neurol. 47, 822–826. Ito, K., Skinkle, K.L., Hicks, T.P., 1999. Age-dependent, steroidspecific effects of oestrogen on long-term potentiation in rat hippocampal slices. J. Physiol. 515 (Pt 1), 209–220.

M. Okada et al. / Epilepsy Research 53 (2003) 81–94 Kawata, Y., Okada, M., Murakami, T., Kamata, A., Zhu, G., Kaneko, S., 2001. Pharmacological discrimination between effects of carbamazepine on hippocampal basal Ca2+ - and K+ -evoked serotonin release. Br. J. Pharmacol. 133, 557–567. Lamas, J.A., Selyanoko, A.A., Brown, D.A., 1997. Effects of a cognition-enhancer, linopridine (DuP996), on M-type potassium currents (IK(M) ) and some other voltage- and ligand-gated membrane currents in rat sympathetic neurons. Eur. J. Neurosci. 9, 605–616. Lerche, C., Scherer, C.R., Seebohm, G., Derst, C., Wei, A.D., Busch, A.E., Steinmeyer, K., 2000. Molecular cloning and functional expression of KCNQ5, a potassium channel subunit that may contribute to neuronal M-current diversity. J. Biol. Chem. 275, 22395–22400. Marrion, N.V., 1997. Control of M-current. Annu. Rev. Physiol. 59, 483–504. Meves, H., Schwarz, J.R., Wulfsen, I., 1999. Separation of M-like current and ERG current in NG108-15 cells. Br. J. Pharmacol. 127, 1213–1223. Murakami, T., Okada, M., Kawata, Y., Zhu, G., Kamata, A., Kaneko, S., 2001. Determination of effects of antiepileptic drugs on SNAREs-mediated hippocampal monoamine release using in vivo microdialysis. Br. J. Pharmacol. 134, 507–520. Noda, M., Obana, M., Akaike, N., 1998. Inhibition of M-type K+ current by linopirdine, a neurotransmitter-release enhancer, in NG108-15 neuronal cells and rat cerebral neurons in culture. Brain Res. 794, 274–280. Obrenovitch, T.P., Zilkha, E., 1996. Inhibition of cortical spreading depression by L-701,324 a novel antagonist at the glycine site of the N-methyl-d-aspartate receptor complex. Br. J. Pharmacol. 117, 931–937. Oka, H., Shimono, K., Ogawa, R., Sugihara, H., Taketani, M., 1999. A new planar multielectrode array for extracellular recording: application to hippocampal acute slice. J. Neurosci. Methods 93, 61–67. Okada, M., Kawata, Y., Mizuno, K., Wada, K., Kondo, T., Kaneko, S., 1998. Interaction between Ca2+ , K+ , carbamazepine and zonisamide on hippocampal extracellular glutamate monitored with a microdialysis electrode. Br. J. Pharmacol. 124, 1277– 1285. Okada, M., Nutt, D.J., Murakami, T., Zhu, G., Kamata, A., Kawata, Y., Kaneko, S., 2001. Adenosine receptor subtypes modulate two major functional pathways for hippocampal serotonin release. J. Neurosci. 21, 628–640. Owens, D.F., Boyce, L.H., Davis, M.B., Kriegstein, A.R., 1996. Excitatory GABA responses in embryonic and neonatal cortical slices demonstrated by gramicidin perforated-patch recordings and calcium imaging. J. Neurosci. 16, 6414–6423. Pieniaszek Jr., H.J., Fiske, W.D., Saxton, T.D., Kim, Y.S., Garner, D.M., Xilinas, M., Martz, R., 1995. Single-dose pharmacokinetics, safety, and tolerance of linopirdine (DuP 996) in healthy young adults and elderly volunteers. J. Clin. Pharmacol. 35, 22–30. Plouin, P., 1997. Benign familial neonatal convulsions and benign idiopathic neonatal convulsions. In: Engel Jr., J., Pedley, T.A. (Eds.), Epilepsy: A Comprehensive Textbook. Lippincott-Raven, Philadelphia, pp. 2247–2255.

93

Schnee, M.E., Brown, B.S., 1998. Selectivity of linopirdine (DuP996), a neurotransmitter release enhancer, in blocking voltage-dependent and calcium-activated potassium currents in hippocampal neurons. J. Pharmacol. Exp. Ther. 286, 709–717. Schroeder, B.C., Kubisch, C., Stein, V., Jentsch, T.J., 1998. Moderate loss of function of cyclic-AMP-modulated KCNQ2/ KCNQ3 K+ channels causes epilepsy. Nature 396, 687–690. Schroeder, B.C., Hechenberger, M., Weinreich, F., Kubisch, C., Jentsch, T.J., 2000. KCNQ5, a novel potassium channel broadly expressed in brain, mediates M-type currents. J. Biol. Chem. 275, 24089–24095. Schwake, M., Pusch, M., Kharkovets, T., Jentsch, T.J., 2000. Surface expression and single channel properties of KCNQ2/ KCNQ3, M-type K+ channels involved in epilepsy. J. Biol. Chem. 275, 13343–13348. Seifert, G., Zhou, M., Dietrich, D., Schumacher, T.B., Dybek, A., Weiser, T., Wienrich, M., Wilhelm, D., Steinhauser, C., 2000. Developmental regulation of AMPA-receptor properties in CA1 pyramidal neurons of rat hippocampus. Neuropharmacology 39, 931–942. Selyanko, A.A., Hadley, J.K., Wood, I.C., Abogadie, F.C., Delmas, P., Buckley, N.J., London, B., Brown, D.A., 1998. Two types of K+ channel subunit, Erg1 and KCNQ2/3 contribute to the M-like current in a mammalian neuronal cell. J. Neurosci. 19, 7742–7756. Singh, N.A., Charlier, C., Stauffer, D., DuPont, B.R., Leach, R.J., Melis, R., Ronen, G.M., Bjerre, I., Quattlebaum, T., Murphy, J.V., McHarg, M.L., Gagnon, D., Rosales, T.O., Peiffer, A., Anderson, V.E., Leppert, M., 1998. A novel potassium channel gene, KCNQ2, is mutated in an inherited epilepsy of newborns. Nat. Genet. 18, 25–29. Shimono, K., Brucher, F., Granger, R., Lynch, G., Taketani, M., 2000. Origins and distribution of cholinergically induced beta rhythms in hippocampal slices. J. Neurosci. 20, 8462– 8473. Smith, J.S., Iannotti, C.A., Dargis, P., Christian, E.P., Aiyar, J., 2001. Differential expression of kcnq2 splice variants: implications to M current function during neuronal development. J. Neurosci. 21, 1096–1103. Tam, S.W., Zaczek, R., 1995. Linopirdine. A depolarizationactivated releaser of transmitters for treatment of dementia. Adv. Exp. Med. Biol. 363, 47–56. Tinel, N., Lauritzen, I., Chouabe, C., Lazdunski, M., Borsotto, M., 1998. The KCNQ2 potassium channel: splice variants, functional and developmental expression. Brain localization and comparison with KCNQ3. FEBS Lett. 438, 171– 176. Walker, M.C., Galley, P.T., Errington, M.L., Shorvon, S.D., Jefferys, J.G., 1995. Ascorbate and glutamate release in the rat hippocampus after perforant path stimulation: a “dialysis electrode” study. J. Neurochem. 65, 725–731. Wallace, R.H., Marini, C., Petrou, S., Harkin, L.A., Bowser, D.N., Panchal, R.G., Williams, D.A., Sutherland, G.R., Mulley, J.C., Scheffer, I.E., Berkovic, S.F., 2001. Mutant GABAA receptor ␥2 -subunit in childhood absence epilepsy and febrile seizures. Nat. Genet. 28, 49–52. Wang, H.S., Pan, Z., Shi, W., Brown, B.S., Wymore, R.S., Cohen, I.S., Dixon, J.E., McKinnon, D., 1998. KCNQ2 and

94

M. Okada et al. / Epilepsy Research 53 (2003) 81–94

KCNQ3 potassium channel subunits: molecular correlates of the M-channel. Science 282, 1890–1893. Watanabe, H., Nagata, E., Kosakai, A., Nakamura, M., Yokoyama, M., Tanaka, K., Sasai, H., 2000. Disruption of the epilepsy KCNQ2 gene results in neural hyperexcitability. J. Neurochem. 75, 28–33.

Zhu, G., Okada, M., Murakami, T., Kamata, A., Kawata, Y., Wada, K., Kaneko, S., 2000. Dysfunction of M-channel enhances propagation of neuronal excitability in rat hippocampus monitored by multielectrode dish and microdialysis systems. Neurosci. Lett. 294, 53–57.