113,

EXPERIMENTALNRUROLOGY

249-254

(1991)

Effects of Applied Currents on Epileptiform Bursts in Vitro HANI KAYYALI AND DOMINIQUE Departments

of

Biomedical Engineering

DURAND

and Neuroscience, Case Western Reserve University, Cleveland, Ohio 44106

an increase in the amplitude of population spikes when intracellular current flowed from the dendrites into the soma (potential highest at the dendrites), and a decrease when current flowed from the soma into the dendrites (potential highest at the soma layer) (4, 10). The threshold of neuronal modulation in those studies was caused by electric fields on the order of 10 mv/mm. However, the study of the effects of electric currents in brain slices has been mainly limited to normal neural responses. Recent experiments showed that such modulatory effects are possible on epileptiform activity. Results obtained from hippocampal tissue exposed to penicillin have shown that field potentials generated by epileptiform bursts in the pyramidal region can be decreased considerably by a short pulse applied to the stratum radiatum (8, 25). In these experiments, the pulse was shown to interrupt the process of synchronization that characterizes epileptic seizures, but does not reduce neuronal firing. Other experiments utilizing extracellular sinusoidal currents have found reduction in strongly evoked responses and potentiation of weakly evoked responses in the hippocampus depending on the polarity of the electric field (1). These results, however, were observed long after the initiation of the stimulation, and hence the inhibition mechanisms are not clear. Analysis of the interaction between extracellular applied fields and neuronal response (16,17) suggests that high spatial gradients such as those generated by monopolar electrodes should be very effective at controlling neural firing. In this study, quasi-static DC currents were applied via a monopolar electrode in order to determine the effect of spatially localized electric fields on epileptic-like bursting activity in penicillin-treated hippocampal tissue. An in vitro preparation utilizing extracellular and intracellular techniques was used in order to measure directly the effects of electric fields on transmembrane potentials.

In this study, results show that applied electric currents can be effective to control the neuronal bursting that characterizes epileptic activity. Recordings from the CA1 region of rat hippocampus treated with penicillin show that local inhibition of epileptiform bursts is possible by short anodic current pulses (60 ms duration) applied extracellularly. Inhibition was evidenced by a large reduction (>90%) in the amplitude of field potential. Data collected from 20 slices with moderate field potentials (50-80% of maximum) showed that current needed for complete inhibition was on the order of 42 + 3 PA. Intracellular recordings in CA 1 cells (n = 13) showed that the decrease in field potential amplitude was accompanied by suppression of intracellular neuronal firing caused by somatic hyperpolarization as measured by transmembrane potentials. The resulting hyperpolarization was on the order of 13 mv below resting potential for weakly epileptiform responses (60% of maximum response), and 50 mv below resting potential for strongly epileptiform activity (>50% of maximum response). These results reveal the existence of a stimulation window within which inhibition of neuronal elements can be achieved without simultaneous excitation. Q 1991 Academic Press, Inc.

INTRODUCTION The ability to modulate neuronal responses by electric stimulation has been the subject of many previous studies (2, 3, 5, 6, 12, 15). Transcortical polarizing DC currents (extracellular) have been shown to affect spontaneous and evoked activity in the motor and visual cortices in adult cat encephale isole preparation (5). Similar work performed on the rat cerebral cortex has found that prolonged changes in the level of cortical activity are possible by brief polarizing currents (2). However, the complexity of the neuronal interconnections and anatomy has prevented detailed description of the mechanisms involved. The development of the in vitro brain slice preparation has allowed direct accessto both intraand extracellular measurements, thereby renewing the interest in studying the effect of electric fields on neural tissue. Constant field studies in the in vitro preparation of rat hippocampus and turtle cerebellum have shown

METHODS The experiments were conducted on brain slices, 400 pm thick, excised from the rat hippocampus (Fig. lA), which were perfused with artificial cerebrospinal fluid (ACSF). ACSF constituted of the following (in mM): NaCl, 124; KCl, 3.75; KH,PO,, 1.25; CaCl,, 2.0; MgSO,, 249 All

Copyright 0 1991 rights of reproduction

0014-4886/91 $3.00 by Academic Press, Inc. in any form reserved.

250

KAYYALI

AND

Orthodromic

A

B Orthodromic

current

41 uA 27 uA

5mV

15uA l5ms Control

C

Blocking

current

(uA)

FIG. 1. Suppression of penicillin-induced neuronal activity by applied anodic currents. (A) The hippocampal slice. Epileptiform activity was recorded in the CA1 region of the hippocampus and was triggered by an orthodromic electrode located in the stratum radiaturn with a 150~as current pulse. The blocking electrode was a 50-pmdiameter stainless steel electrode located above the somatic layer close to the recording electrode. The blocking electrode was activated by a constant current generator with a fixed pulse width (40-60 ms) and a variable amplitude. (B) With no current applied at the blocking electrode, the extracellular response (control) is characteristic of the effects of penicillin with multiple population spikes in the orthodromic-evoked response. As the amplitude of the applied current was increased, an extracellular potential proportional to the current was generated by the extracellular tissue resistance but the amplitude and the number of the population spikes was largely decreased. (C) Data collected from 20 slices, with 50-80% level of excitation, show that complete suppression was attained at current levels of 42 + 3 PA.

DURAND

The maximum field potential produced by strong orthodromic stimuli was selected to represent 100% activation level. The amplitude percentage of field potentials when compared to the maximum response was used to determine the level of excitation of the tissue. Percentage of excitability level was determined in every slice, and results from experiments with similar excitation levels were then compared. Applied current pulses were delivered in the vicinity of the somata layer via another tungsten electrode (close to the recording electrodes). The current amplitude was measured through a special ammeter connected in series with the source that generates the applied current. The pulse width of the blocking current was chosen to cover the entire duration of the epileptiform event and was kept constant through the course of the experiment (40-60 ms). The amplitude of the current was varied in the range 5 to 100 FA anodically and 5 to 20 PA cathodically. Inhibition of epileptiform activity was measured during the application of the anodic current and was determined by the percentage reduction of the field potential amplitude relative to that before the application of the current. In all of the above measurements, the field potential amplitude was determined by the amplitude of the second population spike since extra spikes only appeared in the presence of penicillin. Intracellular impalements were performed with electrode micropositioning controller (Burleigh 6000). The data were amplified and filtered (5 KHz extracellular, 10 KHz intracellular) using a high impedance intracellular amplifier (Axoprobe). Experimental data were then collected via a General Purpose Interface Bus (GPIB) from a Tektronix digital oscilloscope (2230) and stored in a computer (IBM-AT) for further analysis. Transmembrane potentials were monitored during the application of the current and were calculated by the difference between simultaneous extracellular and intracellular measurements (u, = uin - u,,J. RESULTS

Extracellular 2.0; NaHCO,, 26; dextrose, 10. Solution temperature was maintained at 35°C by a closed feedback temperature controller. Penicillin (3500 units/ml) was added to the ACSF in order to generate the epileptiform activity. Extracellular and intracellular potentials were measured in the somatic layer of the CA1 pyramidal cell region via glass micro-electrodes (extracellular electrode impedance: l-5 MQ, intracellular electrode impedance: 50-200 MQ). Epileptiform activity was initiated by a short orthodromic pulse (0.15 ms) delivered via a tungsten electrode, 50 pm, located in the stratum radiatum (Fig. 1A).

Field Potential

Extracellular epileptiform activity was recorded in hippocampal slices treated with penicillin and was characterized by an evoked potential with several peaks and valleys, lasting for as long as 40 ms (Fig. lB, control). The negative spikes (population spikes) were generated by the synchronized activity of a large number of neurons. When positive current pulses were applied with the blocking electrode, noticeable increases in the extracellular potential were measured (Fig. 1B). However, the number and amplitude of the population spikes were gradually decreased with increasing levels of the applied current. This effect was consistently observed

APPLIED

A

CURRENTS

ON

70’7uA F-j

r

Blocking

20 mV

current

@A)

C

EPILEPTIFORM

251

BURSTS

in the number and amplitude of the population spikes (Fig. 2C). Furthermore, in contrast to anodic stimulation, cathodic stimulation at the soma generated excitation at the onset of the pulse increasing the total number of extracellular population spikes as shown in Fig. 2C.

Simultaneous extracellular and intracellular recordings were obtained from 13 cells in order to determine the effect of the applied electric fields on single cells. A typical response recorded intracellularly from a CA1 cell treated with penicillin, before the current was applied, is shown in Fig. 3A (control). The intracellular epileptiform activity is characterized by a number of action potentials, indicating the hyperexcitability of the neurons, overriding a depolarizing wave known as the

I

50 mV

L

B

20 mV !20 ms

20 ms

Control

FIG. 2.

Increased excitation by applied currents in penicillin solution. (A) The current amplitude was increased beyond the level of maximal inhibition. Excitation was generated and occurs at the onset of the stimulation pulse and was indicated by an arrow. (B) Since suppression was generated at low currents and excitation at higher currents, there is a stimulation window within which suppression can be generated without simultaneous excitation. This window is illustrated for three slices with an identical level of stimulation (75%). (C) When cathodic currents were applied, the amplitude and the number of population spikes were also increased. Moreover, excitation appears at the beginning of the pulse. These effects were opposite to those obtained by anodic stimulation and suggest that the mechanism of action was a polarization of the membrane by the applied currents.

in all 50 slices tested. Data collected from 20 slices, exhibiting comparable reduced levels of activation (5080%) showed that total suppression was achieved with extracellular current amplitudes of 42 +- 3 @A (Fig. 1C). This depression was generated without simultaneous excitation showing that neurons were not activated and that the blocking current amplitude was subthreshold. However, recordings from experiments using applied currents greater than 83 + 4 /IA showed excitation at the onset of the pulse (see arrow in Fig. 2A) while maintaining inhibition of the orthodromic-evoked epileptiform response. These results show that there exists a stimulation window in which inhibition can be obtained effectively (Fig. 2B). Current amplitudes that fall within this window can generate inhibition without simultaneous excitation. Cathodic currents produced the opposite effect generating further excitation, as evidenced by an increase

COflhol

~-

_-_-----...---__-----_---_____________ Orthodromic pulse

Orthxlromic

pulse

I

FIG. 3. Simultaneous intra- and extracellular recordings during suppression of penicillin-induced neuronal activity. (A) Intracellular recordings taken from the CA1 pyramidal region during the application of the blocking current. The firing pattern was typical of the penicillin-induced activity with several action potentials overriding a depolarizing potential known as a paraoxysmai depolarizing shift (PDS). With the application of the blocking current, a large positive potential arising from the tissue resistance was observed. However, the number of action potentials was reduced with increasing currents until complete block of the neuronal firing. (B) Extracellular recordings obtained with an electrode located very close to the intracellular electrode showing that the suppression of neuronal firing was taking place in a large number of cells thereby producing a large decrease in the amplitude of the population spikes.

252

KAYYALI

AND

DURAND

weakly excited cells (~50% excitation, n = 4) was shown to drop from 20.3 mv for control to 8 mv at complete inhibition. PDS amplitudes in strongly excited cells (>50% excitation, n = 4) were reduced from 28 mv for control to 12 mv at 100% inhibition. DISCUSSION

i.

-50

-40

-30

Transmembrane

-20 po!ential

-10 (mV)

below restmg

FIG. 4. Transmembrane potentials in two groups of cells (weakly excited cells, ~50% level of excitation and strongly excited cells, 2~0% level of excitation). Transmembrane potentials (intracellular minus extracellular potentials) in the strongly excited cells displayed potentials on the order of -50 mv, whereas the weakly excited cells displayed potentials on the order of -13 mv. All potentials were negative from resting potential indicating that the somatic membrane was hyperpolarized. Although it was not possible to measure it, the dendritic membrane must be depolarized by the applied current. However, since the spike generating channels are located in the somatic region, hyperpolarization of that region was sufficient to block neuronal firing.

paroxysmal depola~zation shift (PDS). Results showed that the pattern of neuronal behavior was dramatically altered once the blocking current was applied (Fig. 3A). As the strength of the applied current was increased, the action potentials were suppressed sequentially starting from the last until all action potentials were abolished. This intracellular pattern was correlated with the gradual reduction in the amplitude of the population spikes recorded extracellularly (Fig. 3B). Simultaneous extracellular andintracellular measurements showed that the induced suppression was accompanied by a gradual change in somatic transmembrane potentials (Fig. 4). These potentials were observed to be hyperpolarizing (below resting), and their amplitude further decreased with inhibition. Transmembrane potentials were also shown to depend on the degree of neuronal excitability. Neurons with a greater number of action potentials were more difficult to block and stronger currents were necessary to generate the same suppression that was observed in neurons with fewer spikes. Transmembrane potentials at complete inhibition of weakly excited cells (~50% level of excitation, n = 4) were 13 1 3 mv, while in strongly excited cells (~50% level of excitation, n = 4) the average drop below resting potential was 50 Z+Z5 mv. Intracellular recordings have also revealed a gradual decrease in the PDS amplitude with the applied current, which could not be observed extracellularly. The average PDS peak amplitude for

These results show that quasi-static (pulsed) anodic electric currents applied to the extracellular space of the somata of hippocampal neurons can suppress epileptiform activity. The decrease in epileptiform activity cannot be related to the inhibitory postsynaptic potentials (IPSP), since these potentials are considerably decreased in the presence of penicillin (19, 20). Also, the loss of the viability of the cells could not contribute to the reduction of the neural firing since the neurons returned to their control activity consistently after the termination of the applied current. Recent studies have shown that reduction of epileptic-like activity can be generated by well-timed short current pulses, whereby externally applied currents can desynchronize pools of cells leading to a cancellation of extracellular potential (8, 25). Recordings from those experiments showed that intracellular activity was still present with the action potentials occurring at various phases. Intracellular recordings with quasi-static current pulses, however, showed a suppression of intracellular activity during the applied current and not a shift in firing pattern indicating that neuronal desynchronization can be ruled out. The mechanism underlying the suppression of intracellular activity is attributed to polarization of the neuronal membrane caused by a portion of the extracellular current that crosses the cell. The direction of flow of this current across the neuronal structure determines whether inhibition or excitation will ensue. Since the anodic current was applied extracellularly at the soma, current enters into the cell from the extracellular space at the soma level hyperpolarizing the neuronal membrane. Simultaneous extracellular and intracellular recordings taken at the soma during the application of the current measured transmembrane potentials below resting levels showing that the soma was indeed hyperpolarized. It is not obvious, however, if somatic hyperpolarization alone was responsible for the suppression of epileptiform activity. A portion of the current that hyperpolarizes the cell at the soma is expected to exit back to the extracellular space at the dendrites level causing membrane depolarization at that site (see Fig. 5). The role of dendritic depolarization in reducing evoked cortical responses has been repeatedly emphasized previously (6, 12). Although dendritic impalements were not performed in this study, results indicate that dendritic depolarization was involved in the inhibition of epilepti-

APPLIED Blocklng

electrode

(anodlc

CURRENTS

ON

pulse)

inward current generates somatic hy~erPOlarlzatlon Outward

Current

Outward current generates dendrltlc devolarization

FIG. 6. The flow of intracellular current in the CA1 cell (simplified structure) during the application of anodic applied fields. Extracellular current enters the cell at the soma level and exits at the dendrites level. Current entering neuronal membrane causes somatic hyperpolarization, but current exiting the membrane causes dendritic depolarization. Hyperpolarization at the soma moves the firing threshold away from resting potential and reduces spike generation. Dendritic depolarization reduces postsynaptic potentials and decreases the synaptic drive. The effect of dendritic de~la~zation on synaptic inputs was suggested to explain the reduction in PDS amplitude.

form activity. Intracellul~ recordings that monitored neuronal firing during the application of the current showed that action potential suppression was accompanied by a decrease in the amplitude of the PDS. This effect is most likely produced by dendritic depolarization at the synaptic site since it is expected to reduce the synaptic drive that is necessary for the generation of the PDS (7, 11, 14, 19). Hence, it is suggested that both somatic hyperpolarization and dendritic depolarization contribute to the suppression of neuronal hyperexcitability, but the degree of involvement of each is not yet clear. As the amplitude of the applied current was increased one would expect excitation of the neuronal elements located near the electrode. Moreover, one would expect this excitation to take place at the onset of the applied current pulse. This excitatory effect was indeed observed but at current levels significantly higher than those required for maximum inhibition (63 r.lA compared to 42 PA). This suggests, therefore, that it is possible to inhibit neuronal elements by applied current pulses without generating excitation. Furthermore, the orthodromically evoked epileptiform event was still suppressed even in the presence of excitation at the onset of the pulse suggesting that the neuronal populations excited were different from those that were inhibited. Although the mechanism of the excitatory effect is still unclear and needs further investigation direct activation of axonal processes is possible. Cathodic applied currents reversed the direction of current crossing the neuronal membrane generating ex-

EPILEPTIFORM

253

BURSTS

citation, instead of inhibition. Results showed that excitation starts at current levels as low as 10 .uA (50 ms). Extracellular recordings showed a number of population spikes at the onset of the pulse, which clearly indicates neuronal excitation. Also, field potential amplitude of burst discharges was increased suggesting that cathodic currents facilitate epileptiform activity. The fact that the amplitude of the epileptiform activity was increased during cathodic currents strongly supports the hypothesis that the suppression observed with anodie currents was due to membrane polarization. Hence, results show that the effect of cathodic applied fields are clearly excitatory with excitation occurring at current amplitudes significantly lower than those during anodic applied currents. This indicates that the threshold for excitation during cathodic stimulation is considerably smaller than that during anodic stimulation, as has been previously noted for axonal stimulation (16). A theoretical calculation of the electric fields in the tissue was made to assess the potential gradient present during the application of the current along the neuron. The voltage distribution along a line parallel to the surface of a semi-infinite media of resistivity p generated by a monopolar point source placed on the surface discontinuity is V(n) = pIl(2 r 11-j). (p is the resistivity, x is the distance along the axon, d is the distance between the electrode and the cell, and 1 is the current injected). The electric field is given by E(x) = -dV/dx, and was found to be 75 mv/mm at maximum inhibition (I = 45 PA, d = 200 pm, and r = 125 Qcm). The electric field, however, changes with distance (exhibits peaks and valleys and not constant). Therefore, it is not possible to compare directly these results to those obtained from constant field studies. However, applied currents capable of generating 20% suppression also produce in the tissue extracellular voltages at the soma on the order of +4 mv, comparable to the voltage present during the positive phase of endogenous activity of epileptiform responses generated by many models such as penicillin, extracellular high potassium solution, or even electric stimulation. Although the effect of cathodic endogenous fields in synchronizing neurons has been established previously (22-24), results from this study suggest that anodic electric fields generated from naturally occurring signals could play a si~ificant role in the regulation of neuronal firing. CONCLUSION

Suppression of epileptiform activity was generated at relatively low anodic current amplitudes without simultaneous excitation. As the current amplitude was increased, however, excitation was observed but at much larger current amplitudes. Therefore, these results reveal the existence of a stimulation window within which

254

KAYYALI

AND

a population of hyperexcitable neurons can be inhibited without excitation. The existence of subthreshold extracellular stimulation could prove to be an important tool for the study of the role of field effects in epileptiform activity in the brain and could lead to the development of a method for controlling epileptic neuronal activity.

DURAND

LANDAU, W. M., AND G. H. BISHOP. 1964. Analysis of the form and distribution of evoked cortical potentials under the influence of polarizing currents. J. Neurophysiol. 27: 788-813. 13* PRINCE, D. A. 1969. Electrophysiology of “epileptic neurons”: spike generation. Electroenceph. Clin. Neurophysial. 26: 47612.

r4 15.

REFERENCES 16. 1. BAWIN, S. M., AND M. L. ABU-ASSAL. 1986. Long term effects of sinusoidal extracellular electric fields in penicillin treated rat hippocampal slices. Brain Res. 399: 195-199. 2. BINDMAN, L. J., AND 0. J. LIPPOLD. 1964. The action of brief polarizing currents on the cerebral cortex of the rat (1) during current flow and (2) in the production of long-lasting after-effects. J. Physiol. 172: 369-382. 3. BISHOP, G. H., AND J. L. O’LEARY. 1950. The effects of polarizing currents on cell potentials and their significance in the interpretation of central nervous system activity. EEG Clin. Neurophysial. 2: 401-416. 4. CHAN, C. Y., AND C. NICHOLSON. 1986. Modulation by applied electric fields of purkinje and stellate cell activity in the isolated turtle cerebellum. J. Physial. 3’71: 89-114. 5. C~UTZFELDT, 0. D., AND G. H. FROMM. 1962. Influence of transcortical DC currents on cortical neuronal activity. Exp. Neurol. 6.

7.

18. 19.

20.

Ann. 21.

6: 436-452.

DENNEY, D., AND J. M. BROOKHART. 1962. The effects of applied polarization on evoked electro-cortical waves in the cat. Electraenceph.

17.

Clin. Neurophysiol.

22.

14: 885-897.

DINGLEDINE, R., AND L. GJERSTAD. 1979. Penicillin blocks hippocampal IPSPs unmasking prolonged EPSPs. Brain Res. 168:

23.

205-209. 8. 9.

DUFUND, D. 1986. Electrical stimulation can inhibit synchronized neuronal activity. Brain Res. 332: 139-144. FABER, D. S., AND H. KORN. 1989. Electrical field effects: Their relevance in central neural networks. Physial. Reu. 69(3): 821863.

10. JEFFERYS,J. G. 1981. Influence of electric fields on the excitability of granule cells in guinea-pig hippocampal slices. J. Physiol. 319: 143-152. 11. JOHNSTON, D., AND T. H. BROWN. 1980. Giant synaptic potential hypothesis for epileptiform activity. Science 211: 294-297.

487.

PRINCE, D. A. 1968. The depolarization shift in “epileptic” neurons. Exp. Neurol. 21: 467-485. PURPURA, D. P., AND J. G. MCMURTRY. 1965. Intracellular activities and evoked potential changes during polarization of motor cortex. J. Neurophysial. 28: 166-185. RANCK, J. B. 1975. Which elements are excited in electrical stimulation of mammalian central nervous system: A review. Brain Res. 98: 417-440. RATTAY, F. 1986. Analysis of models for external stimulation of axons. IEEE Trans. Biomed. Eng. 33(10): 974-977. RICHARDSON, T. L., AND R. W. TURNER. 1984. Extracellular fields influence transmembrane potentials and synchronization of hippocampal neuronal activity. Brain Res. 294: 255-262. SCHWARTZKROIN, P. A., AND D. A. PRINCE. 1978. Cellular and field potential properties of epileptogenic hippocampal slices. Brain Res. 14’7: 117-130. SCHWARTZKROIN, P. A., AND D. A. PRINCE. Penicillin-induced epileptiform activity in the hippocampal in vitro preparation.

24.

25.

26.

Neurol.

1: 463-469.

SCHWARTZKROIN, P. A. 1983. Local circuit considerations and intrinsic neuronal properties involved in hyperexcitability and cell synchronization. In Basic Mechanisms of Neuronal Hyperexcitability. (H. H. Jasper and N. M. Van Gelder, Ed.), pp. 75-108. A. R. Liss, New York. SNOW, R. W., AND F. E. DUDEK. 1984. Electrical fields directly contribute to action potential synchronization during convulsant-induced epileptiform bursts. Brain Res. 323: 114-118. TAYLOR, C. P., AND F. E. DUDEK. 1984. Synchronization without active chemical synapses during hippocampal after discharges. J. Neurophysial. 62(l): 143-155. TAYLOR, C. P., AND F. E. DUDEK. 1984. Excitation of hippocampal pyramidal cells by an electrical field effect. J. Neurophysial. 52(l): 126-142. WARMAN, E., AND DUFUND D. 1989. Desynchronization of epileptiform activity by phase resetting. Invited paper. In Proceedings, Eleventh Annual IEEE-EMBS Conference, Seattle. pp. 1286-1287. WONG, R. K. S., AND PRINCE D. A. 1979. Dendritic mechanisms underlying penicillin induced epileptiform activity. Science 204: 1228-1231.