Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

MTH is an ultrashort-lasting barbiturate anesthetic known for its paradoxical ability to activate epileptiform activity in patients with epilepsy (Musella et al., 1971; Wilder, 1971; Aasly et al., 1984; Wyler et al., 1987). The activation effect of MTH is restricted to epileptogenic brain tissue (Hufnagel et al., 1992; Wennberg et al., 1997a). An identical activation of epileptiform activity is observed in surgically isolated blocs of epileptogenic cortex (Wennberg et al., 1997a, 1997b), which suggests that MTH alters local networks by a process mimicking mechanical cortical disconnection. The activating properties of MTH are puzzling because anesthetic barbiturates are known to potentiate GABA-mediated synaptic inhibition (MacDonald and Olsen, 1994), whereby suppression of epileptiform activity might be expected. To date, the cellular actions of MTH on central nervous system neurons have not been detailed. In an attempt to elucidate possible mechanisms underlying this paradoxical phenomenon, we examined the effects of MTH on isolated hippocampus and parietal cerebral cortex, areas known to produce epileptiform activity under a variety of experimental conditions (Dreier and Heinemann, 1991; Perreault and Avoli, 1992; Hoffman and Prince, 1995; Walther et al., 1986). We demonstrate here that perfusion of slices with 10 to 100 µM MTH enhanced the GABAergic synaptic inhibition, without inducing spontaneous epileptiform activity. Part of the present data has appeared in abstract form (Zhang et al., 1997).

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Extracellular and whole-cell recordings in rat brain slices have been described previously (Zhang et al., 1991, 1993, 1997; Shinno et al., 1997). Briefly, male Wistar rats of 22- to 35-day-old were anesthetized with halothane and decapitated immediately. The hemisectioned brain was mounted on an aluminum block and sliced in ice-cold ACSF using a vibrotome (series 1000; Technical Products International, St. Louis, MO). Eight or nine transverse sections of 400-µm thickness were obtained from each half brain. In some experiments, horizontal slices (400-450 µm) that preserve synaptic connections between entorhinal cortex and hippocampus were obtained from the caudal portion of the brain, using the methods previously described (Walther et al., 1986; Dreier and Heinemann, 1991; Jones and Lambert, 1990; Paré et al., 1992; Bear and Lothman, 1993). After slicing, the slices were maintained in an ACSF at room temperature (22-23°C) for at least 1 hr before transfer to the recording chamber. The ACSF contained (in mM) NaCl 125, KCl 2.5, NaH₂PO₄ 1.25, CaCl₂ 2, MgSO₄ 2, NaHCO₃ 26 and glucose 10. The pH value of the ACSF was 7.4 when aerated with 5% CO₂/95% O₂, and the osmolarity was 300 ± 5 mOsmol.

Extracellular and whole-cell recordings were done in a fully submerged chamber at temperature of 32° to 33°C. Humidified 5% CO₂/95% O₂ was applied over the perfusate to ensure a warm, oxygenated local environment. To evoke synaptic responses in the CA1 region, a bipolar tungsten electrode was placed in CA1 stratum radiatum to stimulate Schaffer collateral afferents. To induce synaptic responses in cortical neurons, a patch pipette filled with 150 mM NaCl was placed in the white matter near the recording site. Constant current pulses (0.1-0.2 msec, 10-80 µA) were generated by a Grass stimulator and delivered via an isolation unit (S88; Grass Medical Instruments, Quincy, MA) every 20 to 30 sec.

For the whole-cell recording, the recording patch pipettes were pulled from borosilicate thin wall glass tubes (TW150F-4; World Precision Instruments, Sarasota, FL) using a two-stage Narishige puller (Tokyo, Japan). Our standard patch pipette solution contained 150 mM potassium gluconate, 2 mM HEPES and 100 µM K-EGTA (Fluka, New York, NY). In some experiments, two thirds of the potassium gluconate was replaced with KCl to increase intracellular Cl and to promote GABA_A-mediated miniature IPSCs. The patch pipette solutions had a pH of 7.25 adjusted with KOH and an osmolarity of 280 ± 10 mOsmol. When filled with these patch pipette solutions, the patch pipettes had a resistance of 3 to 4 M. Extracellular recordings of synaptic field potentials were done using a patch pipette filled with 150 mM NaCl.

Signals were recorded via an Axopatch amplifier (200 B; Axon Instruments, Foster City, CA). For whole-cell voltage-clamp recordings, the lowpass Bessel filter was set at 5 KHz, and series resistance compensation was near 80%. Data were acquired, stored and analyzed using PCLAMP software (version 6.3) through an IBM compatible computer. Digitization was performed using a 12-bit A/D interface (DIGIDATA 1200; Axon Instruments). To measure the changes in membrane input conductance, neurons were voltage-clamped at 50 to 55 mV, and constant voltage pulses (30 mV, 300 msec) were applied continuously every 20 to 30 sec before the afferent stimulation (see fig. 4). Spontaneous activities were stored on a digital tape recorder (VR-10B; Instrutech, New York, NY) and analyzed offline. Due to the limitation in the signal/noise ratio, synaptic currents of 10 pA were not included in data analysis.

MTH (brietal sodium; Eli Lilly Canada, Toronto, Ontario) was directly dissolved in the ACSF at desired concentrations. Ionotropic glutamate receptor antagonists CNQX and D-AP5 were purchased from Tocris Cookson (Ballwin, MO). BMI and TTX were obtained from Sigma Chemical (St. Louis, MO). Chemicals for making the patch pipette solution were obtained from Fluka. All internal and external solutions were made with deionized sterile water (pH 5-6; specific resistance, 18.2 M/cm) from a Milli-Q UV Plus system.

Statistical significance was analysed using Student's t test. Mean ± S.E.M. values are given throughout the text.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Effects of MTH on CA1 f-EPSPs. To reveal the effects of MTH on glutamatergic transmission, synaptic responses were recorded extracellularly from the hippocampal CA1 dendritic region after electrical stimulation of Schaffer collateral afferents. In slices perfused with the standard ACSF, these responses manifested a fast, downward waveform terminating in ~15 msec. They were called f-EPSPs (Lynch and Schubert, 1980; Schwartzkroin, 1981) and were fully abolished by perfusion of slices with 10 µM CNQX for 3 to 4 min, a selective glutamate AMPA/kainate receptor antagonist (fig. 1) (Shinno et al., 1997; Zhang et al., 1997). Perfusion of slices with 10, 50 or 100 µM MTH for 4 to 5 min caused no significant change in amplitude and waveform of the CA1 f-EPSPs (fig. 1), nor induction of spontaneous synaptic activity in the CA1, CA3 (n = 11) or deep-layer neocortex (n = 15). Occasionally, axonal spikes with small amplitudes were observed in the CA1 somatic region after the application of MTH (50 or 100 µM). We did not quantify these responses in the present study because of their infrequent appearance.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Effects of MTH on the f-EPSPs recorded from the hippocampal CA1 dendritic region. A, Bar graph summarizes percentage changes in the amplitude of f-EPSPs in slices perfused with 10, 50 or 100 µM MTH or 20 µM CNQX for ~4 min. The control amplitude of f-EPSPs measured before application of MTH or CNQX was 0.92 ± 0.04, 0.89 ± 0.06, 0.94 ± 0.05 or 0.93 ± 0.06 mV, respectively. Numbers of slices examined are indicated in parentheses. Statistical significance was analyzed using the paired t test. *, P  .001. B, Examples of f-EPSPs collected from a slice before, during and after application of 100 µM MTH and in the presence of 20 µM CNQX. Each record represents an averaged response from four measurements. Stimulation artifacts have been truncated for illustration.

Potentiation by MTH of GABA_A-mediated IPSCs. When voltage-clamped at membrane potentials of 50 to 55 mV, individual CA1 pyramidal neurons and cortical neurons displayed a transient inward current followed by outward currents lasting 200 to 400 msec after the afferent stimulation. We have shown previously that these outward currents represent IPSCs mediated by GABA_A receptors because their amplitudes alter accordingly with transmembrane Cl distribution and are blocked by bicuculline methiodide, a selective GABA_A antagonist (Zhang et al., 1991, 1993). Perfusion of slices with 100 µM MTH caused reversible enhancements of these IPSCs in CA1 and cortical neurons, manifested by an increased amplitude and prolonged decay time course (table 1; fig. 2, A and C). Adding 10 µM CNQX and 50 µM D-AP5 (an NMDA glutamate receptor antagonist) to the perfusate did not substantially alter the potentiated IPCSs (n = 3), confirming the potentiation of GABAergic responses by MTH (fig. 2B).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Enhancement of GABAA-mediated IPSCs by MTH. Hippocampal CA1 neurons (A, B) and a neuron in the deep layer of parietal cortex (C) were voltage-clamped at 50 mV, and IPSCs were evoked by constant stimulation every 15 to 20 sec. Averaged responses from 3 or 4 measurements were collected before, during and after applications of MTH (100 µM, for ~4 min). The levels of holding currents were aligned with the base-line control for comparison. Note in B, the potentiated IPSC was not substantially affected after adding to the perfusate ionotropic glutamate receptor antagonists CNQX (10 µM) and APV (50 µM).

View larger version (36K): [in this window] [in a new window]   Fig. 3.   Spontaneous GABAA-mediated synaptic activities induced by MTH. Membrane currents sampled for 20-sec periods were collected from a cortical neuron (A) and CA1 neurons (B) before and after applications of MTH (100 µM, 3-4 min). Both neurons were voltage clamped at 50 mV constantly. GABAA antagonist bicuculline methiodide (BMI, 10 µM) was added to the MTH-containing perfusate in B (bottom trace).

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Increases by MTH in membrane input conductance. CA1 pyramidal neurons (A and C) and a cortical neuron (B) were voltage-clamped at 60 mV, and constant voltage pulses of 30 mV were injected intracellularly through the patch pipette every 20 sec. Averaged responses from 4 measurements were collected before and after application of 100 µM MTH for ~4 min. The levels of holding currents were aligned with the base-line control for comparison. In C, the slice was treated with 20 µM CNQX and 10 µM bicuculline throughout the recording period.

View larger version (27K): [in this window] [in a new window]   Fig. 5.   Effects of MTH on spontaneous GABA release. A, Representative voltage-clamp records were collected from a CA1 neuron before and after the application of 100 µM MTH for 2 min. The neuron was whole-cell dialyzed with 100 mM KCl and held at 60 mV to induce inward miniature IPSCs. Tetrodotoxin (1 µM) was applied throughout the recording period. Note the prolonged time couse of miniature IPSCs in the presence of MTH. Quantitative measurements of these miniature IPSCs were summarized in histograms in B, showing amplitude distribution and occurrence of miniature IPSCs. Measurements were made from a 2-min recording period before (left, 300 events) or after the MTH application (849 events), respectively. Due to the limitation in signal/noise ratio under our recording conditions, events with amplitude 10 pA were not included in data analysis. The Levenberg-Marquardt fits to data were computed using PCLAMP software. The mean amplitude of miniature IPSCs was calculated to be 32 pA in the control (left, single exponential) and 22 and 28 pA after the MTH application (right, double exponential).

MTH induced bicuculline-sensitive outward currents (hyperpolarization). In concurrence with the potentiated IPSCs, both CA1 and cortical neurons displayed outward shifts in holding currents and increases in membrane conductance after the exposure of slices to 100 µM MTH (table 2; fig. 4). Accordingly, both types of neurons showed membrane hyperpolarization by 4 to 8 mV and decreases in membrane resistance when examined in the current-clamp mode. Similar applications of MTH were without effect on the holding current and conductance measurement in CA1 neurons pretreated with 10 µM bicuculline (table 2; fig. 4), suggesting a MTH-induced hyperpolarization via stimulation of GABA_A receptors.

Effects of MTH on spontaneous synaptic activities. To examine the effect of MTH on spontaneously occurring synaptic activities, horizontal brain slices that preserve entorhinal-hippocampal synaptic connections were used (see Materials and Methods). To promote synaptic activity, these slices were perfused with a modified ACSF that contained 4.5 mM KCl and 50 µM 4-AP (Avoli et al., 1996). In slices perfused with the 4-AP-containing ACSF for  0 min, extracellular recordings from the CA3-hilar region revealed spontaneously occurring synaptic events that occurred regularly, with a frequency range of 0.3 to 2 Hz. Simultaneous extracellular and whole-cell recordings (electrode tip distance of 200-400 µm, five pairs) showed that extracellular responses were correlated in a close temporal relationship with periodic discharges of individual neurons, implying coherent excitation (fig. 6A). These spontaneous activities became slower and irregular in the presence of 10 µM bicuculline and vanished during application of 10 µM CNQX (n = 3), suggesting that the balance between glutamatergic excitation and GABAergic inhibition is required to maintain the rhythmic activity (Avoli et al., 1996). Application of 10 µM MTH for 4 to 5 min caused a reduction in the frequency of these spontaneous events from 1.04 ± 0.15 to 0.81 ± 0.21 Hz (n = 10, P = .035), without changing their amplitudes (0.91 ± 0.20 mV and 1.05 ± 0.22 mV measured before and after MTH application) or the temporal coherence between the extracellular and intracellular events (fig. 6C). Individual events often exhibited longer durations and/or multiple peaks after the application of 10 µM MTH (fig. 5B), implying a higher degree of local synchrony. Increasing the concentration of MTH to 100 µM blocked the spontaneous events in seven of seven slices examined.

View larger version (43K): [in this window] [in a new window]   Fig. 6.   Attenuation by MTH of the 4-AP-induced rhythmic activity. Simultaneous extracellular (top) and whole-cell (bottom) recordings were made from the CA3-hilar region of an entorhinal-hippocampal slice. The slice was perfused with a modified ACSF with 4.5 mM KCl and 50 µM 4-AP throughout the recording period. Records were collected before (A) and after (B) application of 10 µM MTH for 5 min. Individual events () are shown in fast sweeps at far right. C, Cross-correlation plots were generated from the initial part of the data shown in A and B. Note a reduction by MTH in the frequency of the rhythmic activities but not their temporal coherence.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The present experiments show that when examined in the standard in vitro condition, treatments of brain slices with 10 to 100 MTH caused no significant change in evoked glutamatergic transmission in the hippocampal CA1 region, nor induction of spontaneous epileptiform activity in hippocampus and neocortex. Considering that at clinical settings, an intravenous injection of 25 to 100 mg of MTH is typically used to promote electroencephalographic spiking (Musella et al., 1971; Wilder, 1971; Wyler et al., 1987; Hufnagel et al., 1992; Wennberg et al., 1997a) and that an estimate of the circulating blood volume may be 4 to 5 liters in adults, the concentrations of MTH we used in vitro may be close to or in the dose range of MTH in clinical settings. Thus, our data suggest that MTH at clinically relevant concentrations does not, by itself, elicit electrophysiologically detectable epileptiform activity in isolated rat brain slices, at least under our experimental conditions.

It is perhaps not surprising that the GABA_A-mediated IPSCs were enhanced by MTH, because barbiturates are known to prolong the openings of GABA_A-gated Cl channels in response to agonists (MacDonald and Olsen, 1994). However, this mechanism does not seem to fully explain the robust increase in IPSC amplitude induced by MTH in this study or by pentobarbital in previous studies (Zhang et al., 1991, 1993). Given that individual postsynaptic GABA_A receptors are likely to be saturated after evoked GABA release from pre-synaptic terminals (Edwards et al., 1990; Mody et al., 1995), it raises a possibility that barbiturates may enhance the strength of GABA synapses presynaptically, in addition to their action on postsynaptic GABA_A receptors. We show here that MTH application promoted the occurrence of spontaneous miniature IPSCs in the presence of TTX and increased the amplitude of the bicuculline-insensitive IPSCs after afferent stimulation. Although further experiments are needed to fully characterize the kinetics of miniature IPSCs, the present data do suggest that barbiturates, including MTH, may act presynaptically via recruiting more active releasing sites and/or promoting synchronized release from multiple synapses, therefore increasing the amplitude of the evoked IPSC.

In parallel to the potentiation of the IPSCs, MTH application also caused a large outward shift in the holding current with increased conductance in both cortical and hippocampal neurons. Moreover, this outward shift was prevented by pretreating slices with bicuculline, suggesting an induction of a GABA_A receptor-mediated hyperpolarization by MTH. The direct activation of GABA_A receptors by barbiturates has been recognized for sometime (see review by MacDonald and Olsen, 1994), and the gating of GABA_A receptors by barbiturates is functionally similar to that produced by the natural agonist GABA (Rho et al., 1996). In recombinant human GABA_A receptors expressed in Xenopus oocytes, the barbiturate activation of GABA_A receptors is influenced by the alpha subunit, with an affinity of 58, 138 or 528 µM for the combination composed of alpha-6 beta-1 gamma-2s, alpha-2 beta-2 gamma-2s or alpha-5 beta-2 gamma-2s subunits, respectively (Thompson et al., 1996). We propose that the MTH-induced hyperpolarization is an important pharmacological action of this barbiturate anesthetic and that this is likely due to a direct stimulation of GABA_A receptors by MTH. However, further experiments are required to test whether the hyperpolarization induced by MTH may result from tonic release of GABA and subsequent activation of GABA_A receptors.

Induction of synchronous, rhythmic activities in brain slices by low micromolar concentrations of 4-AP is considered to be a useful in vitro model of epileptiform activities (Avoli, 1996). Although the ionic mechanisms by which 4-AP acts are not fully understood, it is generally thought that an elevation of extracellular K⁺, a GABA_A receptor-mediated depolarization and network synchronization all likely play important roles in generating the spontaneous field responses (Michelson and Wong, 1991, 1994; Perreault and Avoli, 1992; Louvel et al., 1994; Avoli et al., 1996; Lamsa and Kaila, 1997). In the present experiments, applications of MTH attenuated (at 10 µM) or abolished (at 100 µM) the 4-AP-induced rhythmic events in the CA3-hilar or CA1 region, probably via enhancing the GABAergic synaptic inhibition and/or imposing the tonic hyperpolarization.

In summary, we demonstrate that MTH enhanced GABAergic inhibition by potentiating GABA_A-mediated IPSCs and inducing a tonic hyperpolarization but did not by itself elicit epileptiform activity in rat brain slices. Based on these in vitro results, we propose that MTH activation of epileptiform activity in clinical settings may be mediated through increased GABA_A-mediated synchronization (Benardo and Wong, 1995; Avoli, 1996; Freund and Buzsáki, 1996) in the setting of preexistent proepileptic networks in epileptogenic brain tissue.