close

Вход

Забыли?

вход по аккаунту

?

Bumetanide enhances phenobarbital efficacy in a neonatal seizure model.

код для вставкиСкачать
Bumetanide Enhances Phenobarbital Efficacy
in a Neonatal Seizure Model
Volodymyr I. Dzhala, PhD,1,2 Audrey C. Brumback, PhD,2 and Kevin J. Staley, MD1,2
Objectives: High levels of expression of the Na⫹-K⫹-2Cl⫺ (NKCC1) cotransporter in immature neurons cause the accumulation of intracellular chloride and, therefore, a depolarized Cl⫺ equilibrium potential (ECl). This results in the outward flux of
Cl⫺ through GABAA channels, the opposite direction compared with mature neurons, in which GABAA receptor activation is
inhibitory because Cl⫺ flows into the cell. This outward flow of Cl⫺ in neonatal neurons is excitatory and contributes to a
greater seizure propensity and poor electroencephalographic response to GABAergic anticonvulsants such as phenobarbital and
benzodiazepines. Blocking the NKCC1 transporter with bumetanide prevents outward Cl⫺ flux and causes a more negative
GABA equilibrium potential (EGABA) in immature neurons. We therefore tested whether bumetanide enhances the anticonvulsant action of phenobarbital in the neonatal brain.
Methods: Recurrent seizures were induced in the intact hippocampal preparation in vitro by continuous 5-hour exposure to
low-Mg2⫹ solution. The anticonvulsant efficacy of phenobarbital, bumetanide, and the combination of these drugs was studied.
Results: Phenobarbital failed to abolish or depress recurrent seizures in 70% of hippocampi. In contrast, phenobarbital in
combination with bumetanide abolished seizures in 70% of hippocampi and significantly reduced the frequency, duration, and
power of seizures in the remaining 30%.
Interpretation: Thus, alteration of Cl⫺ transport by bumetanide enables the anticonvulsant action of phenobarbital in immature brain. This is a mechanistic demonstration of rational anticonvulsant polypharmacy. The combination of these agents may
comprise an effective therapy for early-life seizures.
Ann Neurol 2008;63:222–235
GABA is the major inhibitory neurotransmitter in the
adult mammalian brain. Fast inhibitory action of
GABA is mediated by ionotropic GABAA receptors
(GABAA-Rs). Binding of GABA to the GABAA-Rs
opens channels permeable to chloride and bicarbonate.
Intracellular chloride levels are maintained at a low
level by the electroneutral K⫹-Cl⫺ cotransporter
(KCC2). Thus, GABAA-R activation results in a net
influx of negatively charged chloride in the cell, which
hyperpolarizes the membrane1 and reduces action potential generation. The GABAA-R–mediated anion
conductance also effectively reduces the membrane resistance and thereby shunts excitatory glutamatergic inputs, resulting in further suppression of neuronal activity.2 When GABAA receptors in the hippocampus and
neocortex are blocked, positive feedback mediated by
recurrent excitatory circuits triggers a chain reaction of
neuronal firing, culminating in epileptiform activity.3,4
In contrast with the adult brain, immature neurons
actively accumulate chloride via the electroneutral
Na⫹-K⫹-2Cl⫺ (NKCC1) transporter. Under these
conditions, GABAA-R activation results in a net efflux
of negatively charged chloride ions, which depolarizes
the membrane. The size of the chloride flux is important, and smaller anion effluxes may not trigger action
potentials. However, if the membrane is depolarized
sufficiently to trigger action potentials and open
voltage-gated calcium channels,5–7 GABA action is
clearly excitatory.8,9 Under both conditions, GABAA-R
activation may still be inhibitory by virtue of the
shunting of more strongly depolarizing glutamatemediated activity.10,11 The importance of the shunting
effect of GABA is well established by the finding that
when all GABAA-R are blocked, the net effect is proconvulsant in the neonatal brain.12,13 Thus, synaptically released GABA has a dual action, both excitatory
and inhibitory, in the immature nervous system.14 –18
The dual action of GABA in the developing nervous
system has led to confusion regarding its role in seizures. GABAA-R agonists may increase the frequency
of seizure in the hippocampal network of neonatal rats
in vitro, whereas GABAA-R antagonists reduce or abolish seizures.19,20 On the other hand, greater concentration of the GABAA-R agonists leads to anion fluxes
From the 1Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Boston, MA; and 2Departments of
Neurology and Pediatrics, University of Colorado at Denver and
Health Sciences Center, Denver, CO.
Published online Oct 4, 2007 in Wiley InterScience
(www.interscience.wiley.com). DOI: 10.1002/ana.21229
Received Feb 25, 2007, and in revised form Jul 10. Accepted for
publication Aug 3, 2007.
222
Address correspondence to Dr Staley, Department of Neurology,
Massachusetts General Hospital, 55 Fruit Street, VBK 910, Boston,
MA 02114. E-mail: kstaley@partners.org
© 2007 American Neurological Association
Published by Wiley-Liss, Inc., through Wiley Subscription Services
that overwhelm the transport capacity of NKCC1. Under these conditions, the driving force for chloride efflux is reduced and the predominant effect of
GABAA-R activation is shunting, producing an anticonvulsant effect.19,21,22
Currently, the first-line medical treatment for neonatal seizures is composed of drugs that increase
GABAA-R channel chloride currents: barbiturates and
benzodiazepines.23,24 Although these drugs are effective
anticonvulsants in the more mature brain, the excitatory effect of GABA in immature neurons renders
these anticonvulsants ineffective in controlling the electroencephalographic seizures in the majority of neonatal patients, causing a major therapeutic void in this
vulnerable age group.25–29 Recent experimental data
have linked this reduced efficacy of GABA-enhancing
anticonvulsants to neuronal chloride transport in the
developing brain.19,20,30 Fortunately, the NKCC1 cotransporter is exquisitely sensitive to the diuretic bumetanide.31,32 Bumetanide prevents the accumulation
of Cl⫺ in neurons and reduces or reverses the depolarization (and excitatory) action of GABA,30,33,34 which
attenuates epileptiform activity in neonatal rats.30
Bumetanide has been extensively tested in human
term and preterm infants.35,36 The developmental patterns of NKCC1 and KCC2 expression are similar in
the human and rat cortex.30 It may be reasonable to
combine bumetanide, which blocks the excitatory effect of GABA in immature neurons, with barbiturates,
which enhance GABAA-R channel chloride currents to
treat the neonatal seizures. Blocking or reducing the
GABA-mediated depolarizing action will remove the
excitatory effects of GABA, so that increasing the
GABA-mediated conductance will only serve to increase shunting inhibition. This strategy maximizes the
anticonvulsant efficacy of the GABA system. The combination of these agents would represent an example of
the frequently discussed idea of rational anticonvulsant
polypharmacy. We tested the efficacy of the diuretic
bumetanide in combination with the GABA-enhancing
anticonvulsant phenobarbital in the treatment of recurrent tonic-clonic epileptiform activity in the intact immature hippocampus in vitro.
Materials and Methods
All animal use protocols conformed to the guidelines of the
National Institutes of Health, the University of Colorado
Health Sciences Center animal care and use committee, and
the Massachusetts General Hospital Center for Comparative
Medicine on the use of laboratory animals.
Experimental Systems
Hippocampal slices were prepared from Sprague–Dawley rat
male pups (postnatal days 4 to 7 [P4-7]). Rats were anesthetized and decapitated. Brains were rapidly removed to oxygenated (95% O2-5% CO2), ice-cold (2–5°C) artificial cere-
brospinal fluid (ACSF) containing 126mM NaCl, 3.5mM
KCl, 2mM CaCl2, 1.3mM MgCl2, 25mM NaHCO3,
1.2mM NaH2PO4, and 11mM glucose (pH 7.4). The transverse hippocampal slices (thickness, 400 and 450␮m) were
cut using a Leica VT-1000E vibratome (Leica Microsystems
GmbH, Nussloch, Germany). Slices were incubated in oxygenated ACSF at room temperature (20 –22°C) for at least 1
hour before use. For electrophysiological recordings, slices
were placed into a conventional submerged-type chamber
and continuously superfused with oxygenated ACSF at 32°C
and at a flow rate of 2ml/min. For some experiments, 4-(2hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)–
buffered ACSF was used to prevent HCO3⫺ flux through
the GABAA channel and make GABA equilibrium potential
(EGABA) almost equal to ECl. HEPES-buffered ACSF contained 147.5mM NaCl, 2.5mM KCl, 26mM HEPES, 2mM
CaCl2, 2mM MgCl2, 1.25mM NaH2PO4, and 10mM glucose; pH 7.4 at 32°C.
The intact hippocampal formations were prepared from
P4 to P6 Sprague–Dawley rat male pups. Rats were anesthetized and decapitated. The brain was rapidly removed to oxygenated (95% O2-5% CO2), ice-cold (2–5°C) ACSF containing 126mM NaCl, 3.5mM KCl, 2mM CaCl2, 1.3mM
MgCl2, 25mM NaHCO3, 1.2mM NaH2PO4, and 11mM
glucose (pH 7.4). The hemispheres were separated, and after
removing the cerebellum, frontal part of the neocortex, and
surrounding structures, the intact hippocampi were dissected
from the septohippocampal complex.37 The hippocampi
were incubated in oxygenated ACSF at room temperature
(20 –22°C) for at least 1 hour before use. For recordings, the
hippocampi were placed into a conventional submerged
chamber and continuously superfused with oxygenated
ACSF at 32°C and at a flow rate of 4ml/min. After 10 minutes, the control ACSF was substituted with an ACSF without MgCl2 added (low-Mg2⫹ ACSF). In this nominally
Mg2⫹-free solution, the extracellular concentration of Mg2⫹
was determined by the contamination by Mg2⫹ of the other
components of the ACSF.38
Electrophysiological Recordings and Data Analysis
Extracellular field potentials were recorded in the hippocampal slices and in the intact hippocampal preparations in vitro
using tungsten microelectrodes and a low-noise multichannel
amplifier (band-pass 0.1Hz to 4kHz; ⫻1,000). Microelectrodes made from coated tungsten wire of 50␮m diameter
(California Fine Wire Company, Grover Beach, CA) were
used for simultaneous recordings of population field activity
in electroencephalogram band (0.1–100Hz), fast ripple oscillations (200 –500Hz), and multiunit activity (MUA; 500Hz
high-pass filter). Root mean square noise level with an electrode placed in the perfusion solution was typically 3 to
4␮V, whereas the amplitude of action potentials recorded
from the pyramidal cell layer ranged from this noise level up
to 200␮V. The signals were digitized using an analogue-todigital converter (DigiData 1322A; Axon Instruments, Burlingame, CA). Sampling interval per signal was 100 microseconds (10kHz). pCLAMP 9.2 (Axon Instruments), Mini
Analysis 6.03 (Synaptosoft, Decatur, CA), and Origin 7.5
SR6 (Microcal Software, Northampton, MA) programs were
used for the acquisition and data analysis. Power spectrum
Dzhala et al: Neonatal Seizure Therapy
223
analysis was performed after applying a Hamming window
function. Power was calculated by integrating the root mean
square value of the signal in frequency bands from 1 to
1,000Hz.
Patch-clamp recordings were made using MultiClamp
700A amplifier (Axon Instruments). Differential interference
contrast microscopy (Axioskop FS; Carl Zeiss, Jena, Germany) through a 40⫻ water immersion objective was used
for visual control of experiments. Patch electrodes were made
from borosilicate glass capillaries (G150T-10; Warner Instrument Corporation, Hamden, CT) using the model PP-83H
two-stage micropipette puller (Narishige, Tokyo, Japan).
Patch pipette solution for gramicidin perforated-patch recordings contained 150mM KCl, 10mM HEPES (pH 7.2
adjusted with potassium hydroxide, 290mOsM). Gramicidin
was first dissolved in dimethylsulfoxide to prepare a stock
solution of 40mg/ml and then diluted to a concentration of
80␮g/ml in the patch solution. The gramicidin-containing
solution was prepared daily and sonicated before the experiments. Patch electrodes were back filled with the gramicidin
solution, then tip filled with gramicidin-free solution by applying suction to the back end of the electrode while the tip
was immersed in gramicidin-free solution. The electrode resistance ranged from 4 to 5M⍀. Twenty to 30 minutes after
the cell-attached formation, series resistance decreased and
stabilized at 12 to 30M⍀. The resting membrane potential
(Vm) was measured from control recordings in currentclamp mode with no current injected. A liquid-junction potential of ⫺3.7mV was calculated with the junction potential
calculator in Clampex 8.2 (Axon Instruments) and subtracted from all voltage measurements. The GABAA receptor–mediated currents were pharmacologically isolated by application of 10␮M NBQX, 50␮M D-AP5(D-(⫺)-2-amino5-phosphonopentanoic
acid
(D-APV),
and
1␮M
CGP35348, and measured in voltage-clamp mode. To measure the reversal potentials for the GABA-induced current
(EGABA), we applied 5mV voltage steps, and at each holding
potential (Vh), GABA (100␮M) was pressure applied (Picospritzer-II; Parker Hannifin Corporation, Cleveland, OH)
for 10 to 20 milliseconds at 0.05Hz through a pipette to the
dendrites of the recorded neurons. Drugs were applied for 15
to 20 minutes, and data were collected 10 minutes after
starting drug applications. The change in EGABA and driving
force for GABA action (EGABA-Vm) were measured at that
time. Peak current responses or charge transfer for each voltage command were plotted, and the data were fitted using
Origin 7.5 SR6 software (Microcal Software).
Statistical Analysis
Group measures are expressed as means ⫾ standard error of
the mean; error bars also indicate standard error of the mean.
The statistical significance of differences was assessed with
the Student’s t test. The level of significance was set at p ⬍
0.05.
Chemicals
Reagents were purchased from Sigma-Aldrich (St. Louis,
MO) and Tocris Cookson (Ellisville, MO), prepared as stock
solutions, and stored before use as aliquots in tightly sealed
224
Annals of Neurology
Vol 63
No 2
February 2008
vials at the manufacturers’ recommended temperatures and
conditions.
Results
Effects of Phenobarbital and Phenobarbital in
Combination with Bumetanide on Excitatory Action
of GABA
In the neonatal hippocampus, the excitatory action of
GABA increases spontaneous firing rates of neurons
and contributes to the generation of giant depolarizing potentials (GDPs).16,19,20,30,39 – 41 Spontaneous
action potentials and bursts of network-driven, highfrequency action potentials representing field GDPs
from multiple cells were invariably detected in simultaneous extracellular field potential recordings from
the CA3 and proximal CA1 pyramidal cell layer in P5
to P7 rat hippocampal slices (Fig 1A). Bath application of phenobarbital (100␮M) did not affect the occurrence of GDPs. The mean GDP frequency in control ACSF and 10 to 15 minutes after phenobarbital
application was 6.2 ⫾ 1.4 and 7 ⫾ 2.1 GDPs/min,
respectively (n ⫽ 5; p ⫽ 0.76). Bath application of
bumetanide (10␮M) in the presence of phenobarbital
rapidly depressed synchronous bursts of action potentials, consistent with previously published data regarding the inhibitory effect of bumetanide alone on
spontaneous GDPs.30,41
To determine the effects of phenobarbital and phenobarbital in combination with bumetanide on excitatory actions of GABA, we measured spontaneous firing
rates of CA3 and CA1 hippocampal neurons using
metal electrodes that detect nearby action potentials;
this activity is referred to as MUA. MUA was measured from the same populations of neurons before and
during activation of the GABAA receptors by isoguvacine (see Figs 1B–E). Bath application (1 minute) of
isoguvacine (10␮M) transiently increased neuronal firing rate in the CA3 and CA1 pyramidal cell layer by
288 ⫾ 67% (n ⫽ 7; p ⫽ 0.0038) and 166 ⫾ 35%
( p ⫽ 0.0016), respectively, consistent with an excitatory effect of GABAA receptor activation. In the presence of phenobarbital, the effects of isoguvacine on
spontaneous neuronal firing rate were similar (see Figs
1C, E). Averaged MUA frequency in the CA3 and
CA1 pyramidal cell layer increased by 246 ⫾ 23%
(n ⫽ 5; p ⫽ 0.00003) and 154 ⫾ 13% ( p ⫽
0.00035), correspondingly. In contrast, application of
isoguvacine in the presence of phenobarbital in combination with bumetanide reduced neuronal firing rate in
the CA3 and CA1 pyramidal cell layer by 33 ⫾ 10%
(n ⫽ 6; p ⫽ 0.0029) and 49 ⫾ 9% ( p ⫽ 0.003),
respectively. Thus, bumetanide prevents GABAmediated excitation in cortical neural networks and
may therefore set up a condition for more effective anticonvulsant action of phenobarbital in the neonatal
brain.
pressure application of 100␮M GABA (10-millisecond pulse duration) to the dendrites approximately
100␮m from the soma in CA1 pyramidal cells and
recorded at different holding potentials. Nominally
bicarbonate-free ACSF and pipette solutions were
used to minimize the HCO3⫺ flux through the
GABAA receptor channels. The resting membrane potential of P4 to P6 CA1 pyramidal cells was ⫺70 ⫾
2.1mV (n ⫽ 6). Under control conditions, the mean
EGABA was ⫺63 ⫾ 3.3mV, similar to previously reported data.33,44 In the presence of bath-applied phenobarbital (100␮M), the mean EGABA was ⫺61.3 ⫾
2.6mV, not significantly different from the control
value ( p ⫽ 0.49). These values for EGABA would be
Fig 1. Effects of phenobarbital and phenobarbital in combination with bumetanide on excitatory action of GABA. (A) Extracellular field potential recordings in the CA3 and CA1
pyramidal cell layer in a P6 hippocampal slice. Spontaneous
neuronal activity is characterized by multiunit activity (MUA)
and network-driven bursts of MUA that are also described as
giant depolarizing potentials (GDPs). Phenobarbital (100␮M)
did not block or suppress GDPs (n ⫽ 6). Bath application of
phenobarbital in combination with bumetanide (10␮M) abolished GDPs (n ⫽ 6). (B–D) Effect of the GABAA receptor
agonist isoguvacine (ISO; 10␮M bath applied for 1 minute)
on MUA frequency in control artificial cerebrospinal fluid
(ACSF) (B), in phenobarbital (C), and in phenobarbital in
combination with bumetanide (D). Data from extracellular
field potential recordings in the CA3 and CA1 pyramidal cell
layer in postnatal days 6 and 7 hippocampal slices. (E) Averaged MUA frequency rate in control ACSF and during application of isoguvacine in control ACSF (n ⫽ 7), ACSF ⫹
phenobarbital (n ⫽ 5), or ACSF ⫹ phenobarbital and bumetanide (n ⫽ 6). [Color figure can be viewed in the online
issue, which is available at www.interscience.wiley.com]
Effects of Phenobarbital and Phenobarbital in
Combination with Bumetanide on GABA
Equilibrium Potential
We next determined the cellular mechanisms underlying the eradication of GABA-mediated excitation by
the combination of phenobarbital and bumetanide.
Gramicidin perforated patch-clamp recordings, which
do not alter the intracellular chloride concentration,42,43 were performed in acute hippocampal slices
from P4 to P6 rats. Reversal potentials of GABAevoked currents (EGABA) were measured before and
during phenobarbital and subsequent bumetanide
applications (Fig 2A). Currents were elicited by local
Fig 2. Alteration of Cl⫺ transport by phenobarbital and bumetanide. (A) Measuring GABA equilibrium potential
(EGABA) with gramicidin perforated patch-clamp recordings
from the CA1 pyramidal cell in an acute hippocampal slice
from a postnatal day 5 (P5) rat. Currents were evoked by
pressure application of 100␮M GABA at various holding potentials (Vh) in control, in phenobarbital (100␮M), and phenobarbital in combination with bumetanide (10␮M). (B)
Corresponding charge transfer: holding potential relationships
of the currents evoked by pressure application of GABA. (inset)
Superimposed currents at Vh ⫽ ⫺81mV in control (black
trace), phenobarbital (red trace), and phenobarbital in combination with bumetanide (green trace). Phenobarbital increased the decay time of GABA-evoked currents and/or charge
transfer for GABA-evoked currents. Subsequent application of
bumetanide, in the presence of phenobarbital, caused a negative shift in EGABA. (C) Driving force of GABAA-R–mediated
currents from recordings in control, phenobarbital, and phenobarbital in composition with bumetanide. [Color figure can be
viewed in the online issue, which is available at www.
interscience.wiley.com]
Dzhala et al: Neonatal Seizure Therapy
225
several millivolts more positive in the presence of
physiological bicarbonate concentrations. Phenobarbital increased the decay time of GABA-evoked currents
by 37 ⫾ 6.2% ( p ⫽ 0.004), thereby increasing the
charge transferred by the GABAA receptor channel
chloride currents (see Fig 2B).23 Subsequent bath application of bumetanide (10␮M), a selective inhibitor
of NKCC1,45,46 in presence of phenobarbital, resulted in a significant negative shift of EGABA
(EGABA ⫽ ⫺72.5 ⫾ 2.61mV; p ⫽ 0.018), reducing
the depolarizing driving force for GABAA receptor–
mediated currents (see Fig 2C). The value of EGABA
in the presence of phenobarbital and bumetanide was
not significantly different than the resting membrane
potential, reflecting a passive transmembrane Cl distribution. This Cl⫺ distribution was due to block of
NKCC1 by bumetanide in combination with limited
KCC2 expression at this age. EGABA in bumetanide
was significantly more negative than the mean EGABA
in control conditions and in presence of phenobarbital alone (see Fig 2C). These data indicate that
NKCC1 activity is required for immature pyramidal
cells to maintain increased steady-state intracellular chloride concentration (Cli) such that EGABA is
positive to the resting membrane potential,30,33,34
so that the action of GABA is depolarizing and
excitatory.10,19,20 Pharmacological blocking of the
NKCC1 cotransporter by the diuretic bumetanide alters Cl⫺ transport, shifts EGABA toward more negative
values, and makes possible the anticonvulsant action
of phenobarbital in immature neuronal networks. The
anticonvulsant efficacy of phenobarbital and bumetanide, as well as the combination of these drugs, were
therefore studied in a low-Mg2⫹ model of neonatal
seizures in the intact hippocampal preparations in
vitro.
Low-Mg⫹–Induced Epileptiform Activity in the
Intact Hippocampal Preparations In Vitro
Extracellular field potential recordings of MUA and
population activity were performed in the intact hippocampal preparations in vitro from male Sprague–
Dawley rats at P4 to P6. Spontaneous asynchronous
MUA and a network-driven burst of action potentials
(GDPs) were evident in simultaneous records from the
CA3 pyramidal cell layer along the longitudinal axis
from the septal to temporal pole of the intact hippocampi. Bath application of low-Mg2⫹ ACSF enhanced the excitability of the hippocampal neural network. MUA and GDP frequency increased before the
development of recurrent ictal-like epileptiform activity
(Figs 3 and 4). Onset of ictal-like (seizure) epileptiform
activity was 25 ⫾ 3 minutes (n ⫽ 40 preparations) after
continuous superfusion of low-Mg2⫹ ACSF. Largeamplitude population spikes, high-frequency oscillations
in the gamma and fast ripple frequency ranges
226
Annals of Neurology
Vol 63
No 2
February 2008
(30 –140Hz), and/or barrages of high-frequency
(⬎200Hz) MUA typically characterized the onset of
ictal-like events (see Fig 3). Ictal-like epileptiform activity consisted of paroxysmal tonic and clonic patterns of
population discharges (see Fig 3B). The ictal-tonic pattern lasted 30 to 60 seconds and was characterized by 5
to 12Hz population spikes progressively decreasing in
frequency and increasing in amplitude. The ictal-clonic
pattern lasted 30 to 100 seconds and was characterized
by synchronous network-driven population bursts that
progressively decreased in frequency during the ictal pattern, whereas the intraburst population spikes increased
in both duration and amplitude. The ictal-clonic discharges were followed by a postictal depression with reduced neuronal activity. Similar patterns of epileptiform
activity have been observed in the intact corticohippocampal formations in vitro prepared from the neonatal rats,47– 49 mice,50 and in a kainic acid–induced
model of seizures in neonatal rats in vivo.30
Compared with the initial seizures, recurrent seizures
occurred at progressively shorter interseizure intervals
(ISIs), with progressively smaller durations and higher
amplitudes (see Figs 4A–C, 8B, and 8C). Starting from
onset, the mean ISI significantly decreased by 23% from
25.8 ⫾ 1.9 to 19.9 ⫾ 1.2 minutes (n ⫽ 10; p ⫽ 0.013)
during the first hour of low-Mg2⫹ ACSF application.
During 5 hours of low-Mg2⫹ ACSF application, the
mean ISI significantly decreased by 34% to 17.1 ⫾ 1.6
minutes ( p ⫽ 0.01; see Fig 4C). Simultaneously, the
mean duration of seizures gradually decreased by 21%
from 2.11 ⫾ 0.19 to 1.62 ⫾ 0.1 minutes (see Fig 4C;
p ⫽ 0.035). The amplitude of population spikes during
recurrent tonic-clonic seizures gradually increased (see
Fig 4B). Over a 5-hour period of low-Mg2⫹ ACSF application, the mean power of ictal-like events significantly increased by 95 ⫾ 18% ( p ⫽ 0.00005; see Fig
4C). Thus, the low-Mg2⫹–induced seizures are progressive for their frequency, amplitude, and power.
Low Efficiency of Phenobarbital in the Treatment of
Low-Mg2⫹–Induced Neonatal Seizures
The barbiturates exert anticonvulsive effects by allosteric
modulation of GABAA receptors, increasing the duration
of Cl⫺ channel opening.23,51 Although these compounds are effective antiepileptic drugs in the adult
brain, the excitatory effects of GABA in immature neurons16,19,20 contribute to the low efficacy of these
drugs,30,52 such as occurs in the treatment of neonatal
seizures.26,27
We examined the effects of phenobarbital on lowMg2⫹ ACSF–induced recurrent seizures in the intact
hippocampus of P4 to P6 rats. Phenobarbital (100␮M)
was bath applied after five to eight recurrent seizures
for 120 minutes, a time period providing sufficient
sampling of seizures based on the mean ISI in control
(Figs 5A, C). Phenobarbital completely abolished ictal-
Fig 3. Low-Mg2⫹–induced ictal-like epileptiform activity in the intact hippocampus in vitro from a postnatal day 5 (P5) rat. (A)
Extracellular field potential recordings in the CA3 pyramidal cell layer from the septal and temporal poles of the intact hippocampus. Superfusion with low-Mg2⫹ artificial cerebrospinal fluid (ACSF)–induced ictal-like epileptiform activity consisting of ictal-tonic
and ictal-clonic discharges and postictal depression. (B) Examples of the ictal-tonic and ictal-clonic discharges on an expanded time
scale. Wide-band recordings from 0.1Hz to 5kHz.
like epileptiform activities in 3 of 10 experiments
(30%). In 7 of 10 experiments (70%) phenobarbital
failed to stop recurrent seizures (see Figs 5 and 8).
Over a 2-hour period of phenobarbital application, the
mean ISI increased by 74% from 15 ⫾ 0.6 to 26.1 ⫾
5.3 minutes (n ⫽ 7/10; p ⫽ 0.058). However, the duration of seizure and amplitude of the population
spikes during tonic-clonic discharges increased (see Figs
5B, C). The mean duration of recurrent seizures in the
presence of phenobarbital gradually increased by
11.7% from 1.7 ⫾ 0.11 to 1.9 ⫾ 0.1 minutes (see Figs
5C and 8D; p ⫽ 0.016). The mean power of recurrent
seizures significantly increased by 41.3 ⫾ 15% ( p ⫽
0.036), consistent with a low anticonvulsant efficacy of
phenobarbital in neonatal seizures (see Figs 5B, C).
The frequency of recurrent seizures increased immediately after washing out of phenobarbital.
Alteration of Cl⫺ Transport for Early-Life
Seizure Therapy
Recent data provide evidence that NKCC1 transporter
facilitates seizures in the developing brain.30 The
NKCC1 blocker bumetanide shifted ECl negative in immature neurons30,33,34 and suppressed high-potassium–
induced ictal-like epileptiform activity in hippocampal
slices in vitro.30 The efficacy of bumetanide alone and
phenobarbital in combination with bumetanide to suppress low-Mg2⫹ ACSF–induced seizures were studied in
the intact hippocampus of P4 to P6 rats. Bumetanide
(10␮M) was applied for 120 minutes after five to eight
recurrent ictal-like episodes (Fig 6A). In 2 of 10 experiments, bumetanide abolished seizures (ISI ⬎ 120 minutes). In 8 of 10 experiments, the mean ISI significantly
increased by 86% from 16.1 ⫾ 1.4 to 29.9 ⫾ 7 minutes
( p ⫽ 0.024). The mean duration of recurrent seizures in
the presence of bumetanide gradually decreased by
29.1% from 2.13 ⫾ 0.21 to 1.51 ⫾ 0.11 minutes (see
Figs 6C and 8D; p ⫽ 0.022). The mean power of recurrent seizures significantly decreased by 41.3 ⫾ 15%
( p ⫽ 0.036), suggesting a stronger anticonvulsant action
of bumetanide than anticonvulsant action of phenobarbital (see Figs 6B, 6C, and 8).
Phenobarbital (100␮M) in combination with bumetanide (10␮M) abolished low-Mg2⫹ ACSF–induced sei-
Dzhala et al: Neonatal Seizure Therapy
227
Fig 4. Low-Mg2⫹–induced recurrent seizures in the intact hippocampus in vitro. (A) Continuous superfusion of low-Mg2⫹ artificial
cerebrospinal fluid (ACSF) generated recurrent seizures. Extracellular field potential recording in the CA3 pyramidal cell layer from
the temporal pole of intact hippocampal preparation from a postnatal day 5 (P5) rat. (B, top) First and last seizures from the recording in (A) are shown on an expanded time scale. (B, bottom) Corresponding power spectra in the 0.1 to 1,000Hz frequency
band showing increased amplitude of the population activity in the electroencephalographic frequency range. (C) Interseizure intervals (ISI) and duration of seizures gradually decreased after continuous 5-hour superfusion of low-Mg2⫹ ACSF. Simultaneously, the
power of recurrent seizures gradually increased. Summary data are taken from 10 intact hippocampi from P4 to P6 rats. Power of
the first seizure in each recording was considered as 100%.
zures in 7 of 10 intact hippocampi (70%). The frequency of MUA was increased by application of the two
drugs, and local bursts of action potentials were evident
in multichannel recordings from the CA3 pyramidal cell
layer (Fig 7B). In 3 of 10 experiments (30%), phenobarbital in combination with bumetanide significantly
increased the mean ISI by 423% from 16.8 ⫾ 2.8 to
87.8 ⫾ 6.2 minutes ( p ⫽ 0.0005). The mean duration
and power of recurrent seizures gradually decreased by
20% ( p ⫽ 0.03) and 41.3% ( p ⫽ 0.036), respectively,
suggesting a strong anticonvulsant action of phenobarbital in combination with bumetanide (Fig 8).
To compare the efficacy of drugs, we analyzed the
averaged power of extracellular field potential activity
including MUA, interictal and ictal-like patterns, and
the averaged frequency and duration of recurrent seizures in consecutive 1-hour windows in control and in
the presence of drugs (see Figs 8B, C). The data indicate that the effects of the combination of bumetanide
228
Annals of Neurology
Vol 63
No 2
February 2008
and phenobarbital on seizure occurrence, frequency,
duration, and power are more than twice as large as the
effect phenobarbital alone.
Discussion
Enhancement of inhibitory neurotransmission and attenuation of excitatory transmission are basic anticonvulsant mechanisms.53 GABA is the major inhibitory
neurotransmitter of the mammalian brain, activating
Cl⫺ permeable GABAA ionotropic receptors and Gprotein–coupled GABAB metabotropic receptors.54
Conventional anticonvulsant agents such us barbiturates and benzodiazepines exert their actions by allosteric activation of the GABAA receptor, increasing the
duration and/or frequency of the Cl⫺ channel opening.23,51 However, in the developing brain activation
of Cl⫺-permeable GABAA receptors excite many neurons as a result of increased intracellular chloride level
([Cl]i) and a depolarized Cl⫺ equilibrium potential
Fig 5. Low efficiency of phenobarbital in neonatal seizures. (A) Extracellular field potential recording in the CA3 pyramidal cell layer
in the intact hippocampus in vitro of a P5 rat. Continuous application of low-Mg2⫹ artificial cerebrospinal fluid (ACSF) induced
recurrent tonic-clonic seizures. Phenobarbital (100␮M) was applied for 120 minutes to adequately sample intervals between seizures.
(B) Expansion of the seizures and corresponding power spectra show the increase in phenobarbital. (C) Interseizure intervals (ISIs),
duration, and power of seizures before, during, and after phenobarbital application. Phenobarbital abolished the seizures in 3 of 10
hippocampi (ISI ⬎ 120 minutes). In 7 of 10 hippocampi, phenobarbital increased the ISIs and duration and power of recurrent seizures, suggesting its low anticonvulsant effect. Power of the first seizure in each recording was considered as 100%.
(ECl).10,19,30,33,34,44 Depending on the [Cl]i, synaptic
activation of GABAA receptors may result in a net Cl⫺
influx and hyperpolarization (inhibition) of neuronal
membrane potential or in a net Cl⫺ efflux and depolarization (excitation) of the Vm. Excitatory action of
GABA contributes to greater seizure propensity,19,20
supports epileptogenesis in the developing hippocampus,13,55 and underlies low anticonvulsant efficacy of
the barbiturates and benzodiazepines for electroencephalographic seizures in the developing brain.26,29
Prolonged application of phenobarbital (100␮M) failed
to suppress high-K⫹–induced ictal-like activity in hippocampal slices from neonatal rats but depressed epileptiform activity in brain slices from older rats.30 In
this study, we used a low-Mg2⫹ model of neonatal seizures in the intact immature hippocampal formation in
vitro.49 In contrast with the increased potassium model
of neonatal seizures,56,57 this method does not alter the
energy gradient for NKCC1 and KCC2 cationchloride cotransporters. In addition, the intact hippocampal preparation in vitro provides an advantage
over the slice preparation in vitro, preserving longitudinal intrahippocampal connections.37
Long-term application (120 minutes) of phenobarbital (100␮M) had little effect on recurrent seizures induced by low Mg2⫹ in the intact hippocampus in vitro
(see Figs 4 and 8). Our results differ in some respects
from prior studies that utilized relatively short (30
minutes) applications of phenobarbital in the intact
immature corticohippocampal formation in vitro.21
The difference in results is most likely due to the abbreviated applications of anticonvulsants in the prior
study, which did not provide a sufficient sample of the
long (up to 1 hour) and variable ISIs in the low-Mg2⫹
model of neonatal seizures (see Fig 4). In addition,
Quilichini and colleagues21 used older (P7-8) rats of a
Dzhala et al: Neonatal Seizure Therapy
229
Fig 6. Effects of bumetanide on recurrent seizures. (A) Extracellular field potential recording in the CA3 pyramidal cell layer in the
intact hippocampus of a postnatal day 5 (P5) rat. Continuous application of low-Mg2⫹ artificial cerebrospinal fluid (ACSF)–induced recurrent tonic-clonic seizures. Bumetanide (10␮M) was applied for 120-minute period after six recurrent seizures. (B) Examples of seizures before and during application of bumetanide from the recording in (A) are shown on an expanded time scale. Corresponding power spectra showing depression of seizures by bumetanide. (C) Bumetanide increased the interseizure intervals (ISIs)
and reduced duration and power of recurrent seizures. Data are taken from 10 experiments. Power of the first seizure in each recording was considered as 100%.
different strain. Significant interstrain differences in the
rates of rat brain development (eg, see Talos and colleagues58) make it difficult to compare the two results
directly; however, a more mature Cl⫺ transport system
because of age and strain differences likely contributes
to the greater efficacy of GABAergic anticonvulsants in
Quilichini and colleagues’ study.21
The dual action of GABA in immature brain results
in both proconvulsant and anticonvulsant effects that
are not easily distinguished by experimental pharmacological manipulations that block or activate all
GABAA-R indiscriminately. Bath application of pharmacological agonists and antagonists have thus produced a variety of effects depending on the driving
force for GABAA-R–mediated currents, the model of
ictogenesis, and the concentration of the GABAA-R
modulators utilized.10,11,19,20,22 To further complicate
this issue, large GABAA-R–mediated currents can
change the neuronal chloride concentration, so that
bath application of high concentrations of GABA ago-
230
Annals of Neurology
Vol 63
No 2
February 2008
nists can alter the driving force for GABAA-R–mediated currents, causing ECl to move toward resting
membrane potential.
In the central nervous system, the intracellular Cl⫺
concentration is regulated by the cation-chloride cotransporters, and the activity and/or expression of these
transporters are important determinants of the direction
and magnitude of GABA- and glycine receptor–mediated postsynaptic currents.59 Loop-diuretic–sensitive
NKCC and KCC cotransporters are of particular interest for Cl⫺ homeostasis.30,46,59 – 61 Under normal physiological conditions, the NKCC cotransporter isoform 1
mediates ion entry into cells and facilitates the accumulation of Cl⫺, whereas the KCC cotransporter isoform 2
mediates ion exit and promotes Cl⫺ extrusion. Because
these transporters are electroneutral, ion transport does
not affect membrane potential and transport is not affected by membrane potential. The transporters do not
directly utilize ATP to energize ion translocation but instead are driven by Na⫹ and K⫹ gradients developed by
Fig 7. Alteration of Cl⫺ transport by bumetanide enhances efficacy of phenobarbital in suppression of neonatal seizures. (A) Extracellular field potential recording in the CA3 pyramidal cell layer in the intact hippocampus of a postnatal day 5 (P5) rat. Application of low-Mg2⫹ artificial cerebrospinal fluid (ACSF)–induced recurrent seizures. Bath application of phenobarbital (Ph, 100␮M)
in combination with bumetanide (Bum, 10␮M) abolished recurrent seizures. (B, top) The ictal-like activity and extracellular field
potential activity before and during application of drugs are shown on an expanded time scale. (B, bottom) Power spectra of extracellular field potential activity before and during application of drugs. (C) Summary data of the interseizure intervals (ISIs) and
seizure duration and power before and during application of phenobarbital and bumetanide. In 70% of experiments (7/10), seizures were abolished (ISI ⬎ 120 minutes). In 30% (3/10), seizures were slower and depressed.
Na⫹-K⫹-ATPase pump gradients. The diuretics bumetanide (3-(aminosulfonyl)-5-(butylamino)-4-phenoxybenzoic acid) and furosemide (5-minosulfonyl)-4-chloro-2([2-furanylmethyl] amino) benzoic acid) are well-known
inhibitors of these cotranspoters.32,62 Bumetanide has an
approximately 500-fold greater affinity for NKCC1 (Ki
of approximately 0.1␮M) than for KCC2 (Ki of approximately 25–50␮M). Furosemide acts to inhibit both
NKCC1 and KCC2 with about equal potency (inhibition constant, Ki of approximately 25–50␮M). Therefore, low concentration of bumetanide (2–10␮M) can
be used to inhibit NKCC1 without significantly affecting KCC2.59
In neonates, developmentally regulated early expression of NKCC1 and delayed expression of KCC2 cotransporters facilitates the accumulation of Cl⫺ in neurons.30,33,59 The value of EGABA measured with
gramicidin perforated patch recordings is age dependent
and may be either more positive (depolarizing) or negative (hyperpolarizing) to the resting membrane potential
of pyramidal cells.63 During this particular time window
of postnatal development, depolarizing GABA acts in
cooperation with glutamatergic ␣-amino-3-hydroxy-5methyl-4-isoxazolepropionic acid and N-methyl-Daspartate receptors,64 and provides significant excitatory
drive to the proportion of neurons as demonstrated by
noninvasive measurements of neuronal firing rate.19,20,65
In immature neurons in which NKCC1 messenger
RNA was expressed, a high intracellular concentration of
Cl⫺ [Cl⫺]i and corresponding large depolarizing driving
force for GABAA-R–mediated currents were both reduced by pharmacological inhibition of the NKCC1 cotransporter bumetanide.33,34 Bumetanide had no effect
on resting membrane potential in these cells, even
Dzhala et al: Neonatal Seizure Therapy
231
Fig 8. Alteration of Cl⫺ transport for neonatal seizure therapy. (A) Group data of the low-Mg2⫹ artificial cerebrospinal fluid
(ACSF)–induced ictal-like activity in control, phenobarbital (PB, 100␮M), bumetanide (BUM, 10␮M), and combination of phenobarbital (100␮M) and bumetanide (10␮M) (PB⫹BUM). (B–D) Mean power, frequency, and duration of recurrent seizures in
60-minute windows in control recordings (n ⫽ 10), and before, during, and after drug applications (n ⫽ 10 for each case). Black
bar indicates the time window of drug application.
though a shift in EGABA might have been predicted to
alter the reversal potential for tonic GABA-mediated
conductances; this is consistent with an active homeostatic system for maintenance of resting membrane potential. Bumetanide had no effect on EGABA in low[Cl⫺]i neurons, in which NKCC1 expression was not
detected.33 Consistent with the reduction in the depolarization mediated by endogenously released GABA in
immature neurons, low concentrations of bumetanide
prevent GABA-mediated excitation in the CA3 and
CA1 hippocampal network (see Fig 1), and caused a reversible block of physiological patterns of network activity in the immature hippocampus, GDPs in vitro and
hippocampal sharp waves in vivo.30,34,39 We have previously shown that bumetanide depresses high-K⫹–induced epileptiform activity in hippocampal slices in vitro
from neonatal rats but is ineffective in slices from older
animals.30 In this study, we demonstrate that bumetanide depressed but did not abolish low-Mg2⫹–induced
seizures in the intact hippocampal formation in vitro
(see Fig 6). Anticonvulsant effects of bumetanide should
be even greater in the setting of recurrent neonatal seizures, in light of the further positive shift in EGABA that
these seizures induce.13,55
232
Annals of Neurology
Vol 63
No 2
February 2008
Several animal studies demonstrated antiepileptic effects of certain diuretics in in vitro and in vivo models
of epilepsy66 –71 and in humans.70,72 We propose to
use low concentration of diuretic bumetanide to alter
the ion gradients that underlie the excitatory effects of
GABA to improve the therapy of neonatal seizures by
phenobarbital and benzodiazepines. Our data indicate
a high anticonvulsant efficacy of the combination of
phenobarbital and bumetanide (see Figs 7 and 8), and
predict a similar efficacy for the combination of benzodiazepines and bumetanide. Extensive experience
with the bumetanide in neonates35,36 indicates minimal risk to patients, and the clinical availability of
these agents indicates that this strategy could be rapidly
implemented clinically.
The combination of bumetanide and barbiturate
should obviate the need to use such high concentrations of barbiturate that GABA-mediated ion flux
overwhelms NKCC1 transport capacity, which renders GABA more inhibitory in immature neurons.19,49,73 High concentrations of barbiturates have
been associated with significant side effects such as
apoptotic neurodegeneration in the developing
brain74,75 and late cognitive and behavioral impair-
ment.76,77 Bumetanide and low-dose phenobarbital
could potentially cause similar problems by virtue of
more effective reduction of physiological GABAmediated excitation, which is important for neural
circuit formation.8,9 However, only modest effects on
development have been noted after NKCC1 inhibition for periods of time that far exceed the typical
duration of neonatal seizures,78 and these effects must
be weighed against the substantial deleterious effects
of persistent neonatal seizures.79
The origin and mechanisms of some forms of epilepsy in the adult brain shares several features of the
neonatal brain. Depolarizing synaptic GABA responses in human temporal lobe epilepsy correlates
with generation of epileptiform activity in sclerotic
hippocampus,80 and the anomalous expression of
both cation-chloride cotransporters, NKCC1 and
KCC2, may alter Cl⫺ transport81 under these conditions. Thus, reduction of wrong-way cation-chloride
cotransport and restoration of inhibition by GABA
may also underlie the anticonvulsant efficacy of specific diuretics in epilepsies of the mature nervous system.70,72 The combination of diuretics that target
NKCC1 and anticonvulsants that increase the
GABAA receptor–mediated conductance represent an
example of rational anticonvulsant polypharmacy that
may be of significant benefit in the setting of some
forms of intractable seizures.
This work was supported by the NIH (National Institute of Neurological Disorders and Stroke, R01 NS40109, K.J.S.).
References
1. Kaila K. Ionic basis of GABAA receptor channel function in the
nervous system. Prog Neurobiol 1994;42:489 –537.
2. Staley KJ, Mody I. Shunting of excitatory input to dentate
gyrus granule cells by a depolarizing GABAA receptormediated postsynaptic conductance. J Neurophysiol 1992;68:
197–212.
3. Schwartzkroin PA, Prince DA. Penicillin-induced epileptiform
activity in the hippocampal in vitro prepatation. Ann Neurol
1977;1:463– 469.
4. Schwartzkroin PA. Cellular electrophysiology of human epilepsy. Epilepsy Res 1994;17:185–192.
5. Yuste R, Katz LC. Control of postsynaptic Ca2⫹ influx in developing neocortex by excitatory and inhibitory neurotransmitters. Neuron 1991;6:333–344.
6. Lin MH, Takahashi MP, Takahashi Y, et al. Intracellular calcium increase induced by GABA in visual cortex of fetal and
neonatal rats and its disappearance with development. Neurosci
Res 1994;20:85–94.
7. Leinekugel X, Tseeb V, Ben Ari Y, et al. Synaptic GABAA activation induces Ca2⫹ rise in pyramidal cells and interneurons
from rat neonatal hippocampal slices. J Physiol 1995;487(pt 2):
319 –329.
8. LoTurco JJ, Owens DF, Heath MJ, et al. GABA and glutamate
depolarize cortical progenitor cells and inhibit DNA synthesis.
Neuron 1995;15:1287–1298.
9. Ganguly K, Schinder AF, Wong ST, et al. GABA itself promotes the developmental switch of neuronal GABAergic
responses from excitation to inhibition. Cell 2001;105:
521–532.
10. Ben Ari Y. Excitatory actions of gaba during development: the
nature of the nurture. Nat Rev Neurosci 2002;3:728 –739.
11. Cossart R, Bernard C, Ben-Ari Y. Multiple facets of GABAergic
neurons and synapses: multiple fates of GABA signalling in epilepsies. Trends Neurosci 2005;28:108 –115.
12. Khalilov I, Khazipov R, Esclapez M, et al. Bicuculline induces
ictal seizures in the intact hippocampus recorded in vitro. Eur
J Pharmacol 1997;319:R5–R6.
13. Khalilov I, Le Van QM, Gozlan H, et al. Epileptogenic actions
of GABA and fast oscillations in the developing hippocampus.
Neuron 2005;48:787–796.
14. Mueller AL, Taube JS, Schwartzkroin PA. Development of hyperpolarizing inhibitory postsynaptic potentials and hyperpolarizing response to gamma-aminobutyric acid in rabbit hippocampus studied in vitro. J Neurosci 1984;4:860 – 867.
15. Chen G, Trombley PQ, van den Pol AN. Excitatory actions of
GABA in developing rat hypothalamic neurones. J Physiol
1996;494(pt 2):451– 464.
16. Khalilov I, Dzhala V, Ben Ari Y, et al. Dual role of GABA
in the neonatal rat hippocampus. Dev Neurosci 1999;21:
310 –319.
17. Psarropoulou C, Descombes S. Differential bicuculline-induced
epileptogenesis in rat neonatal, juvenile and adult CA3 pyramidal neurons in vitro. Brain Res Dev Brain Res 1999;117:
117–120.
18. Wells JE, Porter JT, Agmon A. GABAergic inhibition suppresses paroxysmal network activity in the neonatal rodent hippocampus and neocortex. J Neurosci 2000;20:8822– 8830.
19. Dzhala VI, Staley KJ. Excitatory actions of endogenously released GABA contribute to initiation of ictal epileptiform activity in the developing hippocampus. J Neurosci 2003;23:
1840 –1846.
20. Khazipov R, Khalilov I, Tyzio R, et al. Developmental changes
in GABAergic actions and seizure susceptibility in the rat hippocampus. Eur J Neurosci 2004;19:590 – 600.
21. Quilichini PP, Diabira D, Chiron C, et al. Effects of antiepileptic
drugs on refractory seizures in the intact immature corticohippocampal formation in vitro. Epilepsia 2003;44:1365–1374.
22. Isaev D, Isaeva E, Khazipov R, et al. Shunting and hyperpolarizing GABAergic inhibition in the high-potassium model of ictogenesis in the developing rat hippocampus. Hippocampus
2007;17:210 –219.
23. Twyman RE, Rogers CJ, Macdonald RL. Differential regulation of gamma-aminobutyric acid receptor channels by diazepam and phenobarbital. Ann Neurol 1989;25:213–220.
24. Carmo KB, Barr P. Drug treatment of neonatal seizures by neonatologists and paediatric neurologists. J Paediatr Child Health
2005;41:313–316.
25. Connell J, Oozeer R, de Vries L, et al. Clinical and EEG response to anticonvulsants in neonatal seizures. Arch Dis Child
1989;64:459 – 464.
26. Painter MJ, Scher MS, Stein AD, et al. Phenobarbital compared with phenytoin for the treatment of neonatal seizures.
N Engl J Med 1999;341:485– 489.
27. Scher MS, Alvin J, Gaus L, et al. Uncoupling of EEG-clinical
neonatal seizures after antiepileptic drug use. Pediatr Neurol
2003;28:277–280.
28. Boylan GB, Rennie JM, Chorley G, et al. Second-line anticonvulsant treatment of neonatal seizures: a video-EEG monitoring
study. Neurology 2004;62:486 – 488.
29. Sankar R, Painter MJ. Neonatal seizures: after all these years we
still love what doesn’t work. Neurology 2005;64:776 –777.
Dzhala et al: Neonatal Seizure Therapy
233
30. Dzhala VI, Talos DM, Sdrulla DA, et al. NKCC1 transporter
facilitates seizures in the developing brain. Nat Med 2005;11:
1205–1213.
31. Isenring P, Jacoby SC, Payne JA, et al. Comparison of Na-K-Cl
cotransporters. NKCC1, NKCC2, and the HEK cell Na-L-Cl
cotransporter. J Biol Chem 1998;273:11295–11301.
32. Hannaert P, Alvarez-Guerra M, Pirot D, et al. Rat NKCC2/
NKCC1 cotransporter selectivity for loop diuretic drugs. Naunyn Schmiedebergs Arch Pharmacol 2002;365:193–199.
33. Yamada J, Okabe A, Toyoda H, et al. Cl⫺ uptake promoting
depolarizing GABA actions in immature rat neocortical neurones is mediated by NKCC1. J Physiol 2004;557:
829 – 841.
34. Sipila ST, Schuchmann S, Voipio J, et al. The cation-chloride
cotransporter NKCC1 promotes sharp waves in the neonatal rat
hippocampus. J Physiol 2006;573:765–773.
35. Sullivan JE, Witte MK, Yamashita TS, et al. Pharmacokinetics
of bumetanide in critically ill infants. Clin Pharmacol Ther
1996;60:405– 413.
36. Lopez-Samblas AM, Adams JA, Goldberg RN, et al. The pharmacokinetics of bumetanide in the newborn infant. Biol Neonate 1997;72:265–272.
37. Khalilov I, Esclapez M, Medina I, et al. A novel in vitro
preparation: the intact hippocampal formation. Neuron 1997;
19:743–749.
38. Mody I, Lambert JD, Heinemann U. Low extracellular magnesium induces epileptiform activity and spreading depression in
rat hippocampal slices. J Neurophysiol 1987;57:869 – 888.
39. Ben Ari Y, Cherubini E, Corradetti R, et al. Giant synaptic
potentials in immature rat CA3 hippocampal neurones.
J Physiol 1989;416:303–325.
40. Leinekugel X, Khazipov R, Cannon R, et al. Correlated bursts
of activity in the neonatal hippocampus in vivo. Science 2002;
296:2049 –2052.
41. Sipila ST, Schuchmann S, Voipio J, et al. The cation-chloride
cotransporter NKCC1 promotes sharp waves in the neonatal rat
hippocampus. J Physiol 2006;573:765–773.
42. Kyrozis A, Reichling DB. Perforated-patch recording with
gramicidin avoids artifactual changes in intracellular chloride
concentration. J Neurosci Methods 1995;57:27–35.
43. Ebihara S, Shirato K, Harata N, et al. Gramicidin-perforated
patch recording: GABA response in mammalian neurones
with intact intracellular chloride. J Physiol 1995;484(pt 1):
77– 86.
44. Tyzio R, Cossart R, Khalilov I, et al. Maternal oxytocin triggers
a transient inhibitory switch in GABA signaling in the fetal
brain during delivery. Science 2006;314:1788 –1792.
45. Gillen CM, Brill S, Payne JA, et al. Molecular cloning and
functional expression of the K-Cl cotransporter from rabbit, rat,
and human. A new member of the cation-chloride cotransporter family. J Biol Chem 1996;271:16237–16244.
46. Russell JM. Sodium-potassium-chloride cotransport. Physiol
Rev 2000;80:211–276.
47. Khalilov I, Dzhala V, Medina I, et al. Maturation of kainateinduced epileptiform activities in interconnected intact neonatal limbic structures in vitro. Eur J Neurosci 1999;11:
3468 –3480.
48. Luhmann HJ, Dzhala VI, Ben Ari Y. Generation and propagation of 4-AP-induced epileptiform activity in neonatal intact limbic structures in vitro. Eur J Neurosci 2000;12:
2757–2768.
49. Quilichini PP, Diabira D, Chiron C, et al. Persistent epileptiform activity induced by low Mg2⫹ in intact immature brain
structures. Eur J Neurosci 2002;16:850 – 860.
234
Annals of Neurology
Vol 63
No 2
February 2008
50. Moser J, Kilb W, Werhahn KJ, et al. Early developmental alterations of low-Mg2⫹-induced epileptiform activity in the intact corticohippocampal formation of the newborn mouse in
vitro. Brain Res 2006;1077:170 –177.
51. Macdonald RL, Rogers CJ, Twyman RE. Barbiturate regulation
of kinetic properties of the GABAA receptor channel of mouse
spinal neurones in culture. J Physiol 1989;417:483–500.
52. Staley K. Enhancement of the excitatory actions of GABA by
barbiturates and benzodiazepines. Neurosci Lett 1992;146:
105–107.
53. Kwan P, Sills GJ, Brodie MJ. The mechanisms of action of
commonly used antiepileptic drugs. Pharmacol Ther 2001;90:
21–34.
54. Hill DR, Bowery NG. 3H-baclofen and 3H-GABA bind to
bicuculline-insensitive GABA B sites in rat brain. Nature 1981;
290:149 –152.
55. Khalilov I, Holmes GL, Ben-Ari Y. In vitro formation of a secondary epileptogenic mirror focus by interhippocampal propagation of seizures. Nat Neurosci 2003;6:1079 –1085.
56. Jensen MS, Yaari Y. The relationship between interictal and
ictal paroxysms in an in vitro model of focal hippocampal epilepsy. Ann Neurol 1988;24:591–598.
57. Traynelis SF, Dingledine R. Potassium-induced spontaneous
electrographic seizures in the rat hippocampal slice. J Neurophysiol 1988;59:259 –276.
58. Talos DM, Fishman RE, Park H, et al. Developmental regulation of alpha-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid receptor subunit expression in forebrain and relationship to regional susceptibility to hypoxic/ischemic injury.
I. Rodent cerebral white matter and cortex. J Comp Neurol
2006;497:42– 60.
59. Payne JA, Rivera C, Voipio J, et al. Cation-chloride cotransporters in neuronal communication, development and
trauma. Trends Neurosci 2003;26:199 –206.
60. Delpire E. Cation-chloride cotransporters in neuronal communication. News Physiol Sci 2000;15:309 –312.
61. Gamba G. Molecular physiology and pathophysiology of electroneutral cation-chloride cotransporters. Physiol Rev 2005;85:
423– 493.
62. Isenring P, Forbush B. Ion transport and ligand binding by
the Na-K-Cl cotransporter, structure-function studies. Comp
Biochem Physiol A Mol Integr Physiol 2001;130:487– 497.
63. Tyzio R, Ivanov A, Bernard C, et al. Membrane potential of
CA3 hippocampal pyramidal cells during postnatal development. J Neurophysiol 2003;90:2964 –2972.
64. Ben Ari Y, Khazipov R, Leinekugel X, et al. GABAA, NMDA
and AMPA receptors: a developmentally regulated ‘menage a
trois.’ Trends Neurosci 1997;20:523–529.
65. Cohen I, Miles R. Contributions of intrinsic and synaptic activities to the generation of neuronal discharges in in vitro hippocampus. J Physiol 2000;524(pt 2):485–502.
66. Hochman DW, Baraban SC, Owens JW, et al. Dissociation of
synchronization and excitability in furosemide blockade of epileptiform activity. Science 1995;270:99 –102.
67. Gutschmidt KU, Stenkamp K, Buchheim K, et al. Anticonvulsant actions of furosemide in vitro. Neuroscience 1999;91:
1471–1481.
68. Hochman DW, Schwartzkroin PA. Chloride-cotransport blockade desynchronizes neuronal discharge in the “epileptic” hippocampal slice. J Neurophysiol 2000;83:406 – 417.
69. Reid KH, Guo SZ, Iyer VG. Agents which block potassiumchloride cotransport prevent sound-triggered seizures in postischemic audiogenic seizure-prone rats. Brain Res 2000;864:
134 –137.
70. Hesdorffer DC, Stables JP, Hauser WA, et al. Are certain
diuretics also anticonvulsants? Ann Neurol 2001;50:
458 – 462.
71. Margineanu DG, Klitgaard H. Differential effects of cationchloride co-transport-blocking diuretics in a rat hippocampal
slice model of epilepsy. Epilepsy Res 2006;69:93–99.
72. Haglund MM, Hochman DW. Furosemide and mannitol suppression of epileptic activity in the human brain. J Neurophysiol 2005;94:907–918.
73. Isaev D, Isaeva E, Khazipov R, et al. Anticonvulsant action of
GABA in the high potassium-low magnesium model of ictogenesis in the neonatal rat hippocampus in vivo and in vitro.
J Neurophysiol 2005;94:2987–2992.
74. Bittigau P, Sifringer M, Genz K, et al. Antiepileptic drugs and
apoptotic neurodegeneration in the developing brain. Proc Natl
Acad Sci U S A 2002;99:15089 –15094.
75. Bittigau P, Sifringer M, Ikonomidou C. Antiepileptic drugs and
apoptosis in the developing brain. Ann N Y Acad Sci 2003;993:
103–114.
76. Farwell JR, Lee YJ, Hirtz DG, et al. Phenobarbital for febrile
seizures—effects on intelligence and on seizure recurrence.
N Engl J Med 1990;322:364 –369.
77. Reinisch JM, Sanders SA, Mortensen EL, et al. In utero exposure to phenobarbital and intelligence deficits in adult men.
JAMA 1995;274:1518 –1525.
78. Ge S, Goh ELK, Sailor KA, et al. GABA regulates synaptic
integration of newly generated neurons in the adult brain. Nature 2006;439:589 –593.
79. Holmes GL, Ben Ari Y. Seizures in the developing brain: perhaps not so benign after all. Neuron 1998;21:1231–1234.
80. Cohen I, Navarro V, Clemenceau S, et al. On the origin of
interictal activity in human temporal lobe epilepsy in vitro. Science 2002;298:1418 –1421.
81. Palma E, Amici M, Sobrero F, et al. Anomalous levels of Cl⫺
transporters in the hippocampal subiculum from temporal lobe
epilepsy patients make GABA excitatory. Proc Natl Acad Sci
U S A 2006;103:8465– 8468.
Dzhala et al: Neonatal Seizure Therapy
235
Документ
Категория
Без категории
Просмотров
2
Размер файла
846 Кб
Теги
efficacy, mode, bumetanide, phenobarbital, enhance, seizure, neonatal
1/--страниц
Пожаловаться на содержимое документа