Brain Research, 567 (1991) 241-247 t~) 1992 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/92/$03.50
241
BRES 17258
Suppression of spontaneous epileptiform activity with applied currents M. Nakagawa and D. Durand Department of Biomedical Engineering and Neuroscience, Case Western Reserve University, Cleveland, OH 44106 (U.S.A.)
(Accepted 30 July 1991) Key words: Epileptiform activity; Suppression; Applied current; Hippocampus; In vitro
It has beem well established that both applied and endogenous electric fields can modulate neuronal activity in various preparations. In this paper, we present the effects of applied currents on spontaneous epileptiform activity in the CA1 region of the rat hippocampus. A computer-controlled system was designed to detect the spontaneous abnormal activity and then apply current pulses of programmable amplitude with monopolar electrodes in the stratum pyramidale. The epileptiform activity was generated by subpeffusion of the neural tissue with an elevated potassium artificial cerebrospinal fluid (CSF) solution. Extracellular recordings showed that the interictal bursts could be fully suppressed in 90% of the slices by subthreshold currents with an average amplitude of 12.5 #A. Intracellular recordings showed that the anodic currents generated hyperpolarization of the somatic membrane thereby suppressing neuronal firing. This inhibitory effect of applied current pulses is important for the understanding of electric field effects on abnormal neuronal activity and could be an effective means of preventing the spread of epileptiform activity.
INTRODUCTION
Low-level electrical fields in the central nervous system can directly modulate neuronal activity s and are functionally important 9A°. The effects of these fields are mediated by the conduction currents generated within the tissue producing depolarization or hyperpolarization as they cross cellular membranes. Large scale field effects have been recently uncovered in the hippocampus. For example low calcium preparations have revealed an important role of these fields in the synchronization of neuronal firing in the absence of synaptic activity 12'14'22'24. This synchronization of neuronal firing is typical of epileptiform activity and can lead to the generation of large extracellular potentials 19. The extracellular conduction currents generated by these potentials can be large enough to recruit neighboring cells t2'14'18"20'22'23. Externally applied currents or fields can also influence neural excitability. The neuronal activity of Purkinje 4'5, hippocampal dentate granule 13 and pyramidal L2 cells have been modulated by specific orientation of the electric fields. These fields are applied with large electrodes located on both sides of the tissue and generate a spatially constant electric field. The con-
duction currents associated with this field can produce excitation or inhibition of normal neuronal activity depending on the direction of the field with respect to the dendritic tree. However, little is known about the effect of applied fields and currents on abnormal electric activity such as epilepsy. Previous in vivo experiments have shown that currents applied on the surface of the cortex can modulate the underlying neuronal activity a'6. The lack of simultaneous intra- and extracellular measurements made it difficult to study the mechanisms of these effects. There is also a paucity of information about the effects on localized electric fields. Spatially constant fields can generate m e m b r a n e polarization of neuronal structures 2'4'5'~3 but only where the diameter of those structures changes such as branching points and terminations 25. Localized fields produced by point source electrodes should be much more effective since the second spatial derivative of the extracellular voltage near the electrode can produce large transmembrane currents tT. Preliminary experiments using in vitro preparations indeed suggest that localized fields such as those generated by monopolar electrodes can produce large modulatory effects of the amplitude of population spikes in the hippocampus is.
Correspondence: D. Durand, Department of Biomedical Engineering, 3510 Charles B. Bolton Bid., Case Western Reserve University, Cleveland, OH 44106, U.S.A.
242 In this study, the effects of localized applied fields on spontaneous epileptiform activity are analyzed. Recently developed models of epilepsy using high concentrations of potassium can produce electrographic seizures and interictal-like spontaneously occurring events resembling those observed in humans with epilepsy26. Moreover, these events are associated with long-lasting negative potentials recorded extracellulady. The goals of this paper are to
RESULTS In rat hippocampal slices bathed in an elevated potassium A C S F (10.0 [K+]o), spontaneous interictal bursts were recorded from the CA1 region as previously described 26. The bursts were from 30 to greater than 150 ms
determine whether positive (anodic) applied currents could
in duration and appeared at intervals of 0.5-5 s (Fig. 2). The first 30 ms of the interictal activity was similar in all spikes and consisted of regularly spaced population
suppress these spontaneous potentials and to study the mechanisms underlying the effects of these extracellular
spikes. This activity appeared to be a direct response to the orthodromic input from the CA3 region 26 and was
currents. A n in vitro slice preparation was chosen since it allows direct access to both intracellular and extracellular
called the initial response. In addition to these bursts, some slices displayed prolonged afterdischarges lasting up to hundreds of milliseconds (Fig. 2 B - D ) . While there
potentials. The spontaneous activity was detected using computer algorithms and currents were applied with monopolar electrodes in order to localize the field effects.
MATERIALS AND METHODS The study was performed on Sprague-Dawley rats, 200-300 g. Transverse hippocampal slices, 400 gm thick, were prepared and maintained in a recording chamber as previously described7. The perfusion medium was an artificial cerebrospinal fluid (ACSF) with the following composition (in raM): NaC1 124, KCI 3.75, KI-I2PO4 1.25, MgSO4 1.5, CaCI2.2H20 1.5, NaHCO a 24, dextrose 10, and bubbled with 95% 02-5% CO2. Concentrations were adjusted to produce a similar solution but with a potassium concentration of 10 mM (ref. 26). The temperature was fixed at 35 - 1 °C, and warmed, humidified 95% 02-5% CO2 vapor was maintained over the slices. ExtraceUular potentials were recorded from glass microelectrodes filled with 2 M NaC1 (3-7 Mf~). The signals were amplified and low-pass filtered (2 kHz). The electrode was placed in the CA1 region stratum pyramidale. Intracellular potentials were measured using glass microelectrodes filled with 2 M CKI (100-200 MQ). These signals were amplified and low-pass filtered (10 kHz). Transmembrane measurements were obtained by subtracting the voltages recorded by two, closely located, intracellular and extracellular electrodes. The A/D and D/A converters of an M100 computer interface (Modular Instruments) were used as an interface for an IBM PC/AT (Fig. 1). The extracellular signal was sampled at 10 kHz in real-time, and a programmable current pulse was applied with the DIA when epileptiform activity was detected. A two-step algorithm was designed to detect the onset of the epileptiform activity in order to discriminate population spikes from background noise. The extracellular potential was first monitored until a negative slope larger than a preset derivative threshold was detected. The signal was monitored until the slope was again positive. A programmable amplitude voltage was then applied to the D/A converter and converted into a current with an isolated current generator. The current pulse was applied to the tissue with a tungsten stimulating electrode located approximately 100 #m from the recording electrode in the stratum pyramidale. The suppression of neuronal firing was accomplished by applying an anodic current pulse of specified duration, between 50 and 75 ms. Timing, amplitude and duration of the pulse could be adjusted from the computer. The decrease in neuronal activity recorded extracellularly was quantified by measuring peak-to-peak amplitude of the spikes in the potential waveform. Current levels were delivered with a minimum resolution of 1/~A. The current amplitude was adjusted until no polulation spikes could be observed from the extracellular potentials (full inhibition). Only slices showing stable interietal activity were chosen for the study. Moreover, only those with large spike amplitudes (about 5 mV or greater) were included in order to facilitate the detection of the neuronal activity.
was some degree of variability between interictal burst waveforms within a slice (amplitudes, durations and timing of population spikes), the initial response was fairly consistent. Thirty-two slices from 19 rats met the selection criteria (see Materials and Methods). The detection thresholds were set at the beginning of each experiment and the system was able to detect successfully the spikes with an error rate of 2.3%. The amplitude of the largest population spike during the initial response and the number of spikes within the recording window were used as
i_
,,~
computer to oscilloscope v and PCM-VCR
extracellular recording electrode
} _ _ ~ ~ Vto I converter
intracellular recording electrode
/electrode ~ _
--~¢.~.. \ ~
f
/ /
_..,~
ff-a
_ alveus Z s. oriens
s. rad,atum
CA3
Fig. 1. Block diagram of the experimental set-up. Spontaneous epileptiform activity was generated with high potassium ACSF solutions. The neuronal activity was recorded with standard extracellular recording techniques. The extracellular potentials were filtered, sampled and digitized and a computer algorithm was used to detect the presence of a spontaneous event. The computer was then programmed to generate a current pulse through the Digital to Analog converter of programmable amplitude. The resulting waveform was then stored for off-line analysis.
243 indicators of the activity level of the neural population. The amplitude of the event to be blocked is variable. Fig. 3 shows, however, that most events consisted of spikes with an amplitude around 6 mV. The largest spike amplitudes in each interictal event ranged from 2.0 to 12.0 mV with a mean ( - S.D.) value of 6.26 ( - 2.60) (n = 27) (see Fig. 3B). The number of spikes of an inteilctal burst appearing within a 100 ms time window from the time of detection varied from 4.5 to 15.5 spikes (n = 26) with a mean (+ S.D.) of 8.92 (+- 2.85) (see Fig. 3A). Anodic current pulses applied into the extracellular space caused a step increase in the extracellular potential due to the tissue resistance, but decreased the amplitude of the superimposed epileptiform activity (see Fig. 4). Larger current amplitudes increased the size of the step extracellular potentials, but further decreased the population spike magnitudes until complete suppression of the activity. Larger current amplitudes beyond the level of maximal inhibition sometimes resulted in a subsequent increase in activity. In 94% of the slices (30/32), 100% inhibition of the spontaneous epileptiform activity was generated, as indicated by the population spike amplitudes less than 0.2 mV. The mean minimum current amplitude necessary to fully block all activity during the pulse was 12.5 _+ 3.8 # A (mean --- S.D., n = 30), with induced tissue polarization potentials of 42 - 15.8 mV. Fig. 5A,C shows the distribution of currents and voltages which caused complete suppression. Cumulative distil-
butions indicate that about 16 btA or 60 mV could generate complete inhibition in 80% of the slices (Fig. 5B,D). Intracellular recordings (n = 5) were obtained to measure directly the effect of the applied current on neuronal firing and the results were identical in all cells. With no current applied, the cells exhibited a paroxysmal depolarization shift (PDS), about 25 mV in magnitude (Fig. 6B). When afterdischarges were present, the intracellular potentials remained elevated compared to their baseline potential just prior to the activity. Application of the current pulse generated an increase in the intracellular potentials (Fig. 6D); however, the transmembrane potential (Vin-Vout) indicated that the cells were in fact hyperpolailzed (Fig. 6F). As the current amplitude was
A #
lat
PODu
of
~.on
S p i k e s
Interictal
per
l ' '
E v e n t
I '
U@~ m i ' .H r4
i
'1
h
0 k 13 E
t
i
I
o Number
i of
i
I
l•
S D i K e s
J
P o p u l a t i o n
S p i k e
A m p l i t u O e s
P iI
20%
L_
k
L
-a
' ~m00
1
20~0
Fig. 2. Spontaneous interictal bursts preceding and following the application of an extracellular current pulse. A: extracellular potential showing 5 interictal bursts, the third burst was inhibited with an anodic pulse. The baseline during the application of the pulse is not shown B,C: two bursts preceding the inhibition pulse (expanded time scale). Note the similarity of the first 30 ms of activity (orthodromic response). Later population spikes (after discharges) show less similarity.
ii o Amp
•
I
1
] ..... 1 •
of
L a r g e s t
Sp
~o
ike
ta
(mV)
Fig. 3. Frequency distributions of control activity. A: distribution of the mean amplitude of largest polulation spike during the initial segment of the interictal burst. The amplitudes ranged from 2 to 12 inV. B: distribution of the number of population spikes during the recording window. The mean number ranged from 4 to 16 spikes per burst.
244
50.00 20.5 uA 4000
;
" 17.6 uA
-~
3ooo
t~J
5
{3-
J Ld
241
BRES 17258
Suppression of spontaneous epileptiform activity with applied currents M. Nakagawa and D. Durand Department of Biomedical Engineering and Neuroscience, Case Western Reserve University, Cleveland, OH 44106 (U.S.A.)
(Accepted 30 July 1991) Key words: Epileptiform activity; Suppression; Applied current; Hippocampus; In vitro
It has beem well established that both applied and endogenous electric fields can modulate neuronal activity in various preparations. In this paper, we present the effects of applied currents on spontaneous epileptiform activity in the CA1 region of the rat hippocampus. A computer-controlled system was designed to detect the spontaneous abnormal activity and then apply current pulses of programmable amplitude with monopolar electrodes in the stratum pyramidale. The epileptiform activity was generated by subpeffusion of the neural tissue with an elevated potassium artificial cerebrospinal fluid (CSF) solution. Extracellular recordings showed that the interictal bursts could be fully suppressed in 90% of the slices by subthreshold currents with an average amplitude of 12.5 #A. Intracellular recordings showed that the anodic currents generated hyperpolarization of the somatic membrane thereby suppressing neuronal firing. This inhibitory effect of applied current pulses is important for the understanding of electric field effects on abnormal neuronal activity and could be an effective means of preventing the spread of epileptiform activity.
INTRODUCTION
Low-level electrical fields in the central nervous system can directly modulate neuronal activity s and are functionally important 9A°. The effects of these fields are mediated by the conduction currents generated within the tissue producing depolarization or hyperpolarization as they cross cellular membranes. Large scale field effects have been recently uncovered in the hippocampus. For example low calcium preparations have revealed an important role of these fields in the synchronization of neuronal firing in the absence of synaptic activity 12'14'22'24. This synchronization of neuronal firing is typical of epileptiform activity and can lead to the generation of large extracellular potentials 19. The extracellular conduction currents generated by these potentials can be large enough to recruit neighboring cells t2'14'18"20'22'23. Externally applied currents or fields can also influence neural excitability. The neuronal activity of Purkinje 4'5, hippocampal dentate granule 13 and pyramidal L2 cells have been modulated by specific orientation of the electric fields. These fields are applied with large electrodes located on both sides of the tissue and generate a spatially constant electric field. The con-
duction currents associated with this field can produce excitation or inhibition of normal neuronal activity depending on the direction of the field with respect to the dendritic tree. However, little is known about the effect of applied fields and currents on abnormal electric activity such as epilepsy. Previous in vivo experiments have shown that currents applied on the surface of the cortex can modulate the underlying neuronal activity a'6. The lack of simultaneous intra- and extracellular measurements made it difficult to study the mechanisms of these effects. There is also a paucity of information about the effects on localized electric fields. Spatially constant fields can generate m e m b r a n e polarization of neuronal structures 2'4'5'~3 but only where the diameter of those structures changes such as branching points and terminations 25. Localized fields produced by point source electrodes should be much more effective since the second spatial derivative of the extracellular voltage near the electrode can produce large transmembrane currents tT. Preliminary experiments using in vitro preparations indeed suggest that localized fields such as those generated by monopolar electrodes can produce large modulatory effects of the amplitude of population spikes in the hippocampus is.
Correspondence: D. Durand, Department of Biomedical Engineering, 3510 Charles B. Bolton Bid., Case Western Reserve University, Cleveland, OH 44106, U.S.A.
242 In this study, the effects of localized applied fields on spontaneous epileptiform activity are analyzed. Recently developed models of epilepsy using high concentrations of potassium can produce electrographic seizures and interictal-like spontaneously occurring events resembling those observed in humans with epilepsy26. Moreover, these events are associated with long-lasting negative potentials recorded extracellulady. The goals of this paper are to
RESULTS In rat hippocampal slices bathed in an elevated potassium A C S F (10.0 [K+]o), spontaneous interictal bursts were recorded from the CA1 region as previously described 26. The bursts were from 30 to greater than 150 ms
determine whether positive (anodic) applied currents could
in duration and appeared at intervals of 0.5-5 s (Fig. 2). The first 30 ms of the interictal activity was similar in all spikes and consisted of regularly spaced population
suppress these spontaneous potentials and to study the mechanisms underlying the effects of these extracellular
spikes. This activity appeared to be a direct response to the orthodromic input from the CA3 region 26 and was
currents. A n in vitro slice preparation was chosen since it allows direct access to both intracellular and extracellular
called the initial response. In addition to these bursts, some slices displayed prolonged afterdischarges lasting up to hundreds of milliseconds (Fig. 2 B - D ) . While there
potentials. The spontaneous activity was detected using computer algorithms and currents were applied with monopolar electrodes in order to localize the field effects.
MATERIALS AND METHODS The study was performed on Sprague-Dawley rats, 200-300 g. Transverse hippocampal slices, 400 gm thick, were prepared and maintained in a recording chamber as previously described7. The perfusion medium was an artificial cerebrospinal fluid (ACSF) with the following composition (in raM): NaC1 124, KCI 3.75, KI-I2PO4 1.25, MgSO4 1.5, CaCI2.2H20 1.5, NaHCO a 24, dextrose 10, and bubbled with 95% 02-5% CO2. Concentrations were adjusted to produce a similar solution but with a potassium concentration of 10 mM (ref. 26). The temperature was fixed at 35 - 1 °C, and warmed, humidified 95% 02-5% CO2 vapor was maintained over the slices. ExtraceUular potentials were recorded from glass microelectrodes filled with 2 M NaC1 (3-7 Mf~). The signals were amplified and low-pass filtered (2 kHz). The electrode was placed in the CA1 region stratum pyramidale. Intracellular potentials were measured using glass microelectrodes filled with 2 M CKI (100-200 MQ). These signals were amplified and low-pass filtered (10 kHz). Transmembrane measurements were obtained by subtracting the voltages recorded by two, closely located, intracellular and extracellular electrodes. The A/D and D/A converters of an M100 computer interface (Modular Instruments) were used as an interface for an IBM PC/AT (Fig. 1). The extracellular signal was sampled at 10 kHz in real-time, and a programmable current pulse was applied with the DIA when epileptiform activity was detected. A two-step algorithm was designed to detect the onset of the epileptiform activity in order to discriminate population spikes from background noise. The extracellular potential was first monitored until a negative slope larger than a preset derivative threshold was detected. The signal was monitored until the slope was again positive. A programmable amplitude voltage was then applied to the D/A converter and converted into a current with an isolated current generator. The current pulse was applied to the tissue with a tungsten stimulating electrode located approximately 100 #m from the recording electrode in the stratum pyramidale. The suppression of neuronal firing was accomplished by applying an anodic current pulse of specified duration, between 50 and 75 ms. Timing, amplitude and duration of the pulse could be adjusted from the computer. The decrease in neuronal activity recorded extracellularly was quantified by measuring peak-to-peak amplitude of the spikes in the potential waveform. Current levels were delivered with a minimum resolution of 1/~A. The current amplitude was adjusted until no polulation spikes could be observed from the extracellular potentials (full inhibition). Only slices showing stable interietal activity were chosen for the study. Moreover, only those with large spike amplitudes (about 5 mV or greater) were included in order to facilitate the detection of the neuronal activity.
was some degree of variability between interictal burst waveforms within a slice (amplitudes, durations and timing of population spikes), the initial response was fairly consistent. Thirty-two slices from 19 rats met the selection criteria (see Materials and Methods). The detection thresholds were set at the beginning of each experiment and the system was able to detect successfully the spikes with an error rate of 2.3%. The amplitude of the largest population spike during the initial response and the number of spikes within the recording window were used as
i_
,,~
computer to oscilloscope v and PCM-VCR
extracellular recording electrode
} _ _ ~ ~ Vto I converter
intracellular recording electrode
/electrode ~ _
--~¢.~.. \ ~
f
/ /
_..,~
ff-a
_ alveus Z s. oriens
s. rad,atum
CA3
Fig. 1. Block diagram of the experimental set-up. Spontaneous epileptiform activity was generated with high potassium ACSF solutions. The neuronal activity was recorded with standard extracellular recording techniques. The extracellular potentials were filtered, sampled and digitized and a computer algorithm was used to detect the presence of a spontaneous event. The computer was then programmed to generate a current pulse through the Digital to Analog converter of programmable amplitude. The resulting waveform was then stored for off-line analysis.
243 indicators of the activity level of the neural population. The amplitude of the event to be blocked is variable. Fig. 3 shows, however, that most events consisted of spikes with an amplitude around 6 mV. The largest spike amplitudes in each interictal event ranged from 2.0 to 12.0 mV with a mean ( - S.D.) value of 6.26 ( - 2.60) (n = 27) (see Fig. 3B). The number of spikes of an inteilctal burst appearing within a 100 ms time window from the time of detection varied from 4.5 to 15.5 spikes (n = 26) with a mean (+ S.D.) of 8.92 (+- 2.85) (see Fig. 3A). Anodic current pulses applied into the extracellular space caused a step increase in the extracellular potential due to the tissue resistance, but decreased the amplitude of the superimposed epileptiform activity (see Fig. 4). Larger current amplitudes increased the size of the step extracellular potentials, but further decreased the population spike magnitudes until complete suppression of the activity. Larger current amplitudes beyond the level of maximal inhibition sometimes resulted in a subsequent increase in activity. In 94% of the slices (30/32), 100% inhibition of the spontaneous epileptiform activity was generated, as indicated by the population spike amplitudes less than 0.2 mV. The mean minimum current amplitude necessary to fully block all activity during the pulse was 12.5 _+ 3.8 # A (mean --- S.D., n = 30), with induced tissue polarization potentials of 42 - 15.8 mV. Fig. 5A,C shows the distribution of currents and voltages which caused complete suppression. Cumulative distil-
butions indicate that about 16 btA or 60 mV could generate complete inhibition in 80% of the slices (Fig. 5B,D). Intracellular recordings (n = 5) were obtained to measure directly the effect of the applied current on neuronal firing and the results were identical in all cells. With no current applied, the cells exhibited a paroxysmal depolarization shift (PDS), about 25 mV in magnitude (Fig. 6B). When afterdischarges were present, the intracellular potentials remained elevated compared to their baseline potential just prior to the activity. Application of the current pulse generated an increase in the intracellular potentials (Fig. 6D); however, the transmembrane potential (Vin-Vout) indicated that the cells were in fact hyperpolailzed (Fig. 6F). As the current amplitude was
A #
lat
PODu
of
~.on
S p i k e s
Interictal
per
l ' '
E v e n t
I '
U@~ m i ' .H r4
i
'1
h
0 k 13 E
t
i
I
o Number
i of
i
I
l•
S D i K e s
J
P o p u l a t i o n
S p i k e
A m p l i t u O e s
P iI
20%
L_
k
L
-a
' ~m00
1
20~0
Fig. 2. Spontaneous interictal bursts preceding and following the application of an extracellular current pulse. A: extracellular potential showing 5 interictal bursts, the third burst was inhibited with an anodic pulse. The baseline during the application of the pulse is not shown B,C: two bursts preceding the inhibition pulse (expanded time scale). Note the similarity of the first 30 ms of activity (orthodromic response). Later population spikes (after discharges) show less similarity.
ii o Amp
•
I
1
] ..... 1 •
of
L a r g e s t
Sp
~o
ike
ta
(mV)
Fig. 3. Frequency distributions of control activity. A: distribution of the mean amplitude of largest polulation spike during the initial segment of the interictal burst. The amplitudes ranged from 2 to 12 inV. B: distribution of the number of population spikes during the recording window. The mean number ranged from 4 to 16 spikes per burst.
244
50.00 20.5 uA 4000
;
" 17.6 uA
-~
3ooo
t~J
5
{3-
J Ld
