- Email: [email protected]
Spatio-temporal distribution of epileptiform activity in slices from human neocortex: recordings with voltage-sensitive dyes
Epilepsy Research 32 (1998) 224 – 232
Spatio-temporal distribution of epileptiform activity in slices from human neocortex: recordings with voltage-sensitive dyes B. Albowitz a,b,*, U. Kuhnt a, R. Ko¨hling b, A. Lu¨cke b, H. Straub b, E.-J. Speckmann b,c, I. Tuxhorn d, P. Wolf d, H. Pannek e, F. Oppel e a Abteilung Neurobiologie, Max Planck Institut fu¨r Biophys. Chem., P.O. Box 2841, 37070 Go¨ttingen, Germany Institut fu¨r Physiologie, Westfa¨lische Wilhelms-Uni6ersita¨t Mu¨nster, Robert Koch Str. 27a, 48149 Mu¨nster, Germany c Institut fu¨r Experimentelle Epilepsieforschung, Westfa¨lische Wilhelms-Uni6ersita¨t Mu¨nster, Hu¨fferstr. 68, 48149 Mu¨nster, Germany d Epilepsiezentrum Bethel, 33617 Bielefeld, Germany e Klinik fu¨r Neurochirurgie, Krankenanstalten Gilead, Bethel, 33617 Bielefeld, Germany b
Abstract The spatio-temporal distribution of epileptiform activity was investigated in slices from human temporal neocortex resected during epilepsy surgery. Activity was recorded by use of a voltage-sensitive dye and an optical recording system. Epileptiform activity was induced with 10 mM bicuculline and electrical stimulation of layer I. In 10 slices from six patients investigated, epileptiform activity spread across most of the slice. Largest amplitudes were located in layer II/III. Epileptiform activity was characterized by long-lasting potentials with slow rising phases and a low velocity of spread in the horizontal direction (0.044 m/s). This spatio-temporal pattern of epileptiform activity in human slices was similar to that found previously in neocortical slices from guinea pigs with bicuculline. In four of nine human slices investigated under control bath conditions (in non-epileptogenic medium), the spatio-temporal activity patterns were similar to those of guinea pigs in non-epileptogenic medium. In the remaining five human slices, however, the spread in the horizontal direction was significantly larger (4188 mm) in non-epileptogenic medium than that found in slices from guinea pigs (2171 mm). Activity in human slices showing such ‘wide spread’ in control bath conditions occasionally had characteristic features of epileptiform activity. Further work will have to clarify whether these epileptiform features reflect intrinsic epileptiform properties in human tissue slices. © 1998 Elsevier Science B.V. All rights reserved. Keywords: Spatio-temporal distribution; Epileptiform activity; Human temporal neocortex; Bicuculline; Electrical stimulation
* Corresponding author. E-mail: [email protected] 0920-1211/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII S0920-1211(98)00054-0
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
1. Introduction
2. Methods
A characteristic feature of epileptogenesis is the spread of activity across a large area of the brain. The pattern of spread has been investigated previously in different model epilepsies both in vivo (e.g. Petsche et al., 1974; London et al., 1989) and in the in vitro slice preparation (e.g. Chervin et al., 1988; Chagnac-Amitai and Connors, 1989a,b; Wadman and Gutnick, 1993). In guinea pig neocortical slices, we have shown by use of optical recordings that supragranular cortical layers are significant both for the generation and spread of epileptiform activity (Albowitz et al., 1990; Albowitz and Kuhnt, 1995). Spread was shown to be independent of the process of induction of epileptiform activity (Albowitz and Kuhnt, 1993c), and was characterized by a low velocity of horizontal (i.e. parallel to the cortical lamination) propagation (B0.04 m/s). Characterization of the spread of activity in animal experiments, however, does not necessarily reflect the situation in human epileptic brain tissue. More important than species differences in cortical structure and function is the acute nature of many animal epilepsy models. Chronic seizure disorder (‘epilepsy’) as seen in human patients is possibly accompanied by histopathological changes of the cortex, which might influence the pattern and mechanisms of spread. Therefore, we analyzed the spread of activity in slices made from human tissue resected during epilepsy surgery (Kato et al., 1973; Schwartzkroin and Prince, 1976). Electrophysiology of neurons in human slices was found to be in many aspects similar to those in neurons of other mammals (see e.g. Williamson, 1994). Epileptiform activity can readily be induced in human slices by application of bicuculline (Avoli and Olivier, 1989; Hwa et al., 1992) or by treatment with low-Mg2 + medium (Avoli et al., 1991; Straub et al., 1992). Using bicuculline to activate epileptiform activity, we monitored spread of this activity in slices from human temporal cortex and compared it to spread in guinea pig cortex under comparable experimental conditions.
2.1. Slice preparation
225
Slices were obtained from a small portion of human neocortical tissue which is normally excised for treatment of pharmacoresistant focal epilepsy originating in the temporal lobe. The patients (n= 6) had received a variety of antiepileptic drugs, which included one or more of the following: carbamazepine, phenytoin, valproate, vigabitrine, and gabapentin. Surface EEG recordings and magnetic resonance imaging (MRI) were performed in all patients included in this study, and the excised tissue was investigated for neuropathological features. In all patients, surface EEG recordings showed temporal lobe interictal spikes. In five out of six patients, MRI and pathology indicated hippocampal sclerosis. Three of six patients showed atrophy of the temporal lobe or hypometabolism in positron emission tomography, and two patients had cortical dysplasia. The experiments were approved by the local ethics committee, and informed consent was obtained from all patients. Slices from the inferior temporal gyrus (Fig. 1) were cut on a vibratome (400 mm). They were placed in a portable incubation chamber (Ko¨hling et al., 1995) with oxygenated (O2/CO2, 95%/5%, v/v) artificial cerebrospinal fluid (ACSF) at a temperature of 28°C and pH 7.4. The composition of
Fig. 1. Preparation of neocortical slices from human tissue resected during epilepsy surgery and from guinea pig neocortex.
226
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
the ACSF was (in mM): NaCl (124), KCl (5), NaH2PO4 (1.25), MgSO4 (1.3), CaCl2 (1), NaHCO3 (26), and glucose (10). Slices were allowed to recover for at least 2 h before they were stained with the voltage-sensitive dye and transferred to the recording chamber. In the recording chamber, the concentration of CaCl2 was raised to 2 mM and the temperature to 35°C. The results from 10 slices of human temporal neocortical tissue were compared to our previous results from studies of sensory neocortical slices of guinea pigs (Fig. 1, Albowitz et al., 1990; Albowitz and Kuhnt, 1993a,b,c).
2.2. Stimulation and recording techniques Details on the techniques used for stimulation, and for electrical and optical recording, have been provided elsewhere (Albowitz and Kuhnt, 1993a,b). Briefly, slices were stained with the voltage-sensitive dye RH795. Activity-dependent fluorescence signals were registered with a fast optical recording system based on a 10× 10 photodiode array, providing a spatial resolution of 94 mm/diode and a temporal resolution of 0.4 ms. Optical signals were corrected for dye bleaching and for staining and illumination irregularities. For data analysis, a baseline was calculated for each photodiode record as the average amplitude of the 250 data points (100 ms) preceding the stimulus. The first point after the stimulus that was followed by \100 data points that were above baseline was defined as onset of the stimulus-evoked response. For each photodiode record, a window of 150 ms was set from this onset point. From these window settings, onset latencies and mean window amplitudes were calculated for each individual photodiode record. Epileptiform activity was induced with bath application of 10 mM bicuculline methiodide and single- or double-pulse electrical stimulation of layer I or the white matter. Monopolar stimulation electrodes consisted of tungsten in glass (50 – 200 kV) and had tip diameters below 30 mm. The pulse width was 30 ms. Stimulation strength was twice the intensity needed to evoke detectable optical signals under control conditions.
The position of the photodiode array with respect to the slice and to the stimulation electrodes was determined photographically in situ after the recording procedure and correlated to cortical layers identified in the histological section prepared following the experiments.
3. Results Following electrical stimulation in both nonepileptogenic and epileptogenic medium, a decrease of fluorescence, which occurred with specific spatio-temporal patterns, was observed in all slices. This decrease of fluorescence corresponds linearly to membrane depolarization (Albowitz and Kuhnt, 1993a). Even though optical signals are recorded and analyzed in this study, we keep the electrophysiological terminology such as ‘potentials’ and ‘depolarization’.
3.1. Epileptiform acti6ity The spatio-temporal distribution of activity evoked by electrical stimulation of layer I is shown in Fig. 2. In this example, the distribution of activity under control bath conditions (top) and in 10 mM bicuculline (bottom) is compared for neocortical slices from human temporal lobe and from guinea pig brain. In each case, evoked epileptiform activity in bicuculline was distributed across most of the slice, with largest amplitudes in layer II/III (see also Fig. 4B and Fig. 5B). This general phenomenon was observed in all slices investigated from human (n= 10) and from guinea pig (n= 32, Albowitz et al., 1990; Albowitz and Kuhnt, 1993c) neocortical tissue. Potentials in human neocortical slices in 10 mM bicuculline recorded distal from the stimulation site were of long duration (\280 ms) and had a slow rising phase. When double-pulse stimulation was used, potentials did not show two distinguishable responses (Fig. 3A,B). The velocity of spread of activity in the horizontal direction (parallel to cortical layering) was rapid when measured close to the stimulation site, but significantly slower at more distant sites (Fig. 3B, Fig. 5C). These features were previously used to define epileptiform
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
227
Fig. 2. Distribution of activity in neocortical slices from human and guinea pig under control conditions and with 10 mM bicuculline. The activity following stimulation of layer I is shown in false color coding, with red as greatest and purple as lowest activity. The area of the slice from which records were obtained is indicated by frames in the schematic diagrams at the bottom. Activity was recorded as relative fluorescence change (dF/F) from slices stained with the fluorescence dye RH795. The results shown here reflect activity summated over a window of 150 ms. WM, white matter.
activity (Albowitz and Kuhnt, 1993c). A similar pattern was observed in guinea pig neocortical slices in 10 mM bicuculline (Albowitz and Kuhnt, 1993c). The fast velocity of spread seen close to the stimulation site was similar to that of activity under control conditions, whereas the slow spread in the periphery was a characteristic feature of epileptiform activity. The velocity of this slow spread in human slices was 0.0449 0.007 m/s (mean9 standard error of the mean, n = 9), similar to the value obtained from guinea pig neocortical slices (0.0429 0.004 m/s, n= 10). Epileptiform activity as defined above always
appeared earliest in supragranular layers before invading middle and deep cortical layers, even with the stimulation electrode in the white matter. In the example shown in Fig. 3A and Fig. 4, the velocity of spread in human slices in the vertical direction was 0.198 m/s. This spread is significantly faster than the velocity of spread of epileptiform activity in the horizontal direction, and similar to the vertical spread of activity under control (i.e. non-bicuculline) conditions. In neocortical slices from guinea pigs, spread of activity in the vertical direction was also faster (0.3919 0.030 m/s, n = 10) than spread in the
228
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
horizontal direction. However, the velocity of spread in the vertical direction in slices from human tissue was highly variable. Occasionally, jumps in onset latency of up to 50 ms were observed (e.g. Fig. 4C, right). Because of this high variability, spread in the vertical direction was not systematically investigated. Stimulation of the white matter in human neocortical slices, even when treated with bicuculline, only rarely (three out of 10 cases) evoked epileptiform activity. In those cases where it was evoked, the spatio-temporal distribution of epileptiform activity was the same as that evoked by layer I stimulation. In contrast, in guinea pig neocortical slices, epileptiform activity could readily be evoked by white matter stimulation. The spatio-temporal distribution of activity in guinea pigs was also similar following layer I and white matter stimulation. 3.2. Acti6ity under control bath conditions
Fig. 3. Traces from selected photodiodes (dots in the schematic diagrams at the right) following electrical stimulation of layer I in a human neocortical slice. Records taken under control conditions (light traces) and with bicuculline (heavy traces) are superimposed. Double-pulse stimulation of layer I (50-ms interval), time points of stimulation are indicated by small arrows. (A) Records of activity from photodiodes representing the vertical axis of the slice (layers I–V). (B) Records of activity from photodiodes representing the horizontal axis of the slices (along layer II/III). The position of the stimulation electrode is indicated by the large arrow. Distal from the stimulation site, activity (under both control and epileptogenic conditions) shows a slow rising phase, but does not show a double-pulse response.
During control conditions in human slices, stimulation of layer I evoked activity of lower amplitude and limited spread (Fig. 2). Double-pulse stimulation evoked two clearly distinguishable responses, showing paired pulse facilitation. As reported previously for guinea pig neocortical slices (Albowitz and Kuhnt, 1993a), each response consisted of two components: an early, fast rising part of short duration (possibly resembling non-synaptic activity) and a later slower and longer-lasting component. In the vertical direction, largest responses were recorded in layers I and II. Activity was restricted predominantly to upper cortical layers (Fig. 2, Fig. 4B). The velocity of spread in the vertical direction was similar to the vertical spread of activity under bicuculline conditions, and similar to that seen in guinea pig neocortical slices (Fig. 4C). In the horizontal direction, the range (i.e. distance) of spread along layer III differed among nine human slices investigated (Figs. 5 and 6). Two groups could be identified: five slices with ‘wide spread’ (4188 9 243 mm, n= 5) and four slices with ‘small spread’ (24469 115 mm, n= 4). The range of spread in the latter group did not significantly differ from that seen in guinea pigs (2171986 mm, n= 7).
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
229
Fig. 4. Mean window amplitudes (B) and onset latencies (C) along the vertical axis of a human slice close (left) and far (right) from the stimulation site in layer I. The position of the stimulation electrode, the cortical layers, and positions of photodiodes from which measurements were made are shown in the schematic drawing of the slice (A). The distribution of amplitudes and latencies during control conditions and with bicuculline is compared.
The difference in the range of spread in human slices with ‘small spread’ vs. ‘wide spread’ (1742 mm) was significant (P B0.001, unpaired t-test). In human slices bathed in normal medium, the velocity of spread in the horizontal direction was rapid at sites close to the stimulation electrode; in most cases, the entire area was activated almost simultaneously (Fig. 5C, central area). This was also observed for spread in the horizontal direction in guinea pig neocortical slices. However, in those human slices showing ‘wide spread’ the velocity of horizontal spread in the periphery was
occasionally much slower, and similar to that of epileptiform activity (Fig. 5C, peripheral area). In the example in Fig. 5C, the velocity of horizontal spread in the periphery was 0.031 m/s under control conditions and 0.035 m/s with bicuculline. Also, in human slices showing ‘wide spread’ under control conditions, recordings from the periphery occasionally showed characteristic properties of epileptiform activity: they had a slow rising phase and long duration and appeared as a single response following double-pulse stimulation (marked by arrows in Fig. 3B). However, epilepti-
230
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
form properties were not seen consistently in slices with wide spread. In temporally subsequent recordings from the same site, some records showed epileptiform properties while others did not. Thus, the activity of the slice appeared to ‘switch’ between two conditions.
4. Discussion
Fig. 5. Mean window amplitudes (B) and onset latencies (C) along layer II/III of a human slice following stimulation of layer I. The position of the stimulation electrode, as well as cortical layers and positions of photodiodes from which measurements were made, are shown in the schematic drawing of the slice (A). The distribution of amplitudes and latencies during control conditions and with bicuculline is compared.
The present results show that the spread of epileptiform activity induced by electrical stimulation in medium containing bicuculline, is similar in slices from human neocortex resected during epilepsy surgery and in slices from guinea pig neocortex. In both preparations, horizontal spread occurs across most of the slice, rapidly at sites close to the stimulating electrode but at a low velocity more distally. The amplitude of the response is largest in supragranular layers and it is at this level that the earliest onset latencies are found. Epileptiform potentials had a slow rising phase and were of long duration. Under the slice conditions, differences in structure and function of the cortex due to the specific species or the cortical area investigated (temporal cortex in humans vs. primary sensory cortex in guinea pigs) did not influence the pattern of epileptiform spread. In contrast to animal experiments, however, vertical spread of activity was not consistent in the human slice. This inconsistency may be due to the fact that, because of the strong gyration of the human brain, slices from human temporal lobe could not always be prepared exactly perpendicular to the cortical surface. As a consequence, vertically oriented neuronal connections might have been separated by the slicing procedure. Possibly for the same reason, epileptiform activity could not always be elicited by white matter stimulation. In the guinea pig neocortical slice, supragranular layers play a predominant role in the spread of epileptiform activity (Albowitz and Kuhnt, 1995). After separating supra- from infragranular layers by a horizontal cut, the supragranular portion was sufficient for the generation and spread of bicuculline-induced epileptiform activity. Also, spread was delayed when horizontal connections were separated in supragranular layers, but not if
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
Fig. 6. Range of spread along layer III under control conditions following stimulation of layer I in slices from guinea pigs and from two sets of human slices (means9 S.E.M.).
they were sectioned in infragranular layers. Since epileptiform activity frequently appeared earliest in supragranular layers regardless of the point of stimulation, supragranular structures were not only sufficient, but seemed to be preferred for generation and propagation of epileptiform activity. While lesion experiments have not been performed in the human slice preparation yet, the spatio-temporal pattern of epileptiform activity reported here also points to a significant role of supragranular layers. Epileptiform activity in the human slice always appeared in supragranular layers before infragranular structures were involved. This was also the case for the rare observations when epileptiform activity could be elicited by white matter stimulation. This predominant role of upper cortical layers is in contrast to findings by Connors (1984) and Chagnac-Amitai and Connors (1989b) who ascribe a pivotal role to intrinsically bursting neurons of layer V for the initiation, synchronization, and spread of epileptiform discharge. It remains to be seen whether the involvement of supra- versus infragranular layers in epileptogenesis is dependent on the specific epilepsy model used, e.g. the bicuculline versus the low-Mg2 + model. In contrast to the similarities of the spatio-temporal activity pattern of the bicuculline-induced model epilepsy in human and guinea pig neocortical slices, the activity pattern under control condi-
231
tions differed considerably. Here, features of epileptiform activity (wide spread, low velocity of spread, slow rising phases, and long-lasting potentials) were frequently present in the human but never in the guinea pig slice. Previously, abnormal responses from cortical neurons (Strowbridge et al., 1992), synchronously active neurons (Schwartzkroin and Haglund, 1986), and spontaneously occurring field potentials (Ko¨hling et al., 1995) have been observed in human slices under control conditions and might reflect epileptiform activity. These observations, and the epileptiform features seen in the present study, suggest that human neocortical slices obtained from tissue removed during epilepsy surgery do preserve features of its epileptogenicity in vitro. In the present investigation, the epileptic focus was not necessarily part of the tissue used for slice preparation. Thus, ‘epileptiform’ features of the activity seen in human tissue may reflect the consequences of chronic projection of epileptiform activity from the focus and may reflect intrinsic abnormalities (structural and/or functional) in related cortical tissue. Clarification of this issue will require further studies.
Acknowledgements
The authors would like to thank B. Herrenproth, S. Lausmann, E. Naß, U. Steveling, and A. Tlustochowski for excellent technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft (DFG 268/3 to B.A.).
References Albowitz, B., Kuhnt, U., 1993a. Evoked changes of membrane potential in guinea pig sensory neocortical slices: an analysis with voltage-sensitive dyes and a fast optical recording method. Exp. Brain Res. 93, 213 – 225. Albowitz, B., Kuhnt, U., 1993b. The contribution of intracortical connections to horizontal spread of activity in the neocortex as revealed by voltage sensitive dyes and a fast optical recording method. Eur. J. Neurosci. 5, 1349 – 1359. Albowitz, B., Kuhnt, U., 1993c. Spread of epileptiform potentials in the neocortical slice — recordings with voltage-sensitive-dyes. Brain Res. 631, 329 – 333.
232
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
Albowitz, B., Kuhnt, U., 1995. Epileptiform activity in the guinea pig neocortical slice spreads preferentially along supragranular layers—Recordings with voltage-sensitive dyes. Eur. J. Neurosci. 7, 273–284. Albowitz, B., Kuhut, U., Ehrenreich, L., 1990. Optical recording of epileptiform voltage changes in the neocortical slice. Exp. Brain Res. 81, 241–256. Avoli, M., Olivier, A., 1989. Electrophysiological properties and synaptic responses in the deep layers of the human epileptogenic neocortex in vitro. J. Neurophysiol. 61, 589– 606. Avoli, M., Drapeau, C., Louvel, J., Pumain, R., Olivier, A., Villemure, J.-G., 1991. Epileptiform activity induced by low extracellular magnesium in the human cortex maintained in vitro. Ann. Neurol. 30, 499–596. Chagnac-Amitai, Y., Connors, B.W., 1989a. Horizontal spread of synchronized activity in neocortex and its control by GABA-mediated inhibition. J. Neurophysiol. 61, 747– 758. Chagnac-Amitai, Y., Connors, B.W., 1989b. Synchronized excitation and inhibition driven by intrinsically bursting neurons in neocortex. J. Neurophysiol. 62, 1149–1162. Chervin, R.D., Pierce, P.A., Connors, B.W., 1988. Periodicity and directionality in the propagation of epileptiform discharges across neocortex. J. Neurophysiol. 60, 1695–1713. Connors, B.W., 1984. Initiation of synchronized neuronal bursting in neocortex. Nature 310, 685–687. Hwa, G.G.C., Avoli, M., Olivier, A., Villemure, J.-G., 1992. Bicuculline-induced epileptogenesis in the human neocortex maintained in vitro. Exp. Brain Res. 83, 329–339. Kato, H., Ito, Z., Matsuoko, S., Sukurai, Y., 1973. Electrical activities in neurons in the sliced human cortex in vitro. Electroencephalogr. Clin. Neurophysiol. 35, 457–462. Ko¨hling, R., Lu¨cke, A., Straub, H., Speckmann, E.-J., Tux-
horn, I., Wolf, P., Pannek, H., Oppel, F., 1995. Spontaneously occurring sharp field potentials in human neocortical slice preparation. Pflu¨gers Arch. Eur. J. Physiol. 6 (Suppl.), R53. London, J.A., Cohen, L.B., Wu, J.-Y., 1989. Optical recordings of the cortical response to whisker stimulation before and after the addition of an epileptogenic agent. J. Neurosci. 9, 2182 – 2190. Petsche, H., Prohaska, O., Rappelberger, R., Vollmer, R., Kaiser, A., 1974. Cortical seizure patterns in multidimensional view: The information content of equipotential maps. Epilepsia 15, 439 – 463. Schwartzkroin, P.A., Haglund, M.M., 1986. Spontaneous rhythmic synchronous activity in epileptic human and normal monkey cortex. Epilepsia 27, 523 – 533. Schwartzkroin, P.A., Prince, D.A., 1976. Microphysiology of human cerebral cortex studied in vitro. Brain Res. 115, 497 – 500. Straub, H., Ko¨hling, R., Speckmann, E.-J., 1992. Low magnesium induced epileptiform discharges in neocortical slices (guinea pig): increased antiepileptic efficacy of organic calcium antagonist verapamil with elevation of extracellular K + concentration. Comp. Biochem. Physiol. 103C, 57 – 63. Strowbridge, B.W., Masukawa, L.M., Spencer, D.D., Shepherd, G.M., 1992. Hyperexcitability associated with localizable lesions in epileptic patients. Brain Res. 587, 158 – 163. Wadman, W.J., Gutnick, M.J., 1993. Non-uniform propagation of epileptiform discharge in brain slices of rat neocortex. Neuroscience 52, 255 – 262. Williamson, A., 1994. Electrophysiology of epileptic human neocortical and hippocampal neurons maintained in vitro. Clin. Neurosci. 2, 47 – 52.
.
.
Spatio-temporal distribution of epileptiform activity in slices from human neocortex: recordings with voltage-sensitive dyes B. Albowitz a,b,*, U. Kuhnt a, R. Ko¨hling b, A. Lu¨cke b, H. Straub b, E.-J. Speckmann b,c, I. Tuxhorn d, P. Wolf d, H. Pannek e, F. Oppel e a Abteilung Neurobiologie, Max Planck Institut fu¨r Biophys. Chem., P.O. Box 2841, 37070 Go¨ttingen, Germany Institut fu¨r Physiologie, Westfa¨lische Wilhelms-Uni6ersita¨t Mu¨nster, Robert Koch Str. 27a, 48149 Mu¨nster, Germany c Institut fu¨r Experimentelle Epilepsieforschung, Westfa¨lische Wilhelms-Uni6ersita¨t Mu¨nster, Hu¨fferstr. 68, 48149 Mu¨nster, Germany d Epilepsiezentrum Bethel, 33617 Bielefeld, Germany e Klinik fu¨r Neurochirurgie, Krankenanstalten Gilead, Bethel, 33617 Bielefeld, Germany b
Abstract The spatio-temporal distribution of epileptiform activity was investigated in slices from human temporal neocortex resected during epilepsy surgery. Activity was recorded by use of a voltage-sensitive dye and an optical recording system. Epileptiform activity was induced with 10 mM bicuculline and electrical stimulation of layer I. In 10 slices from six patients investigated, epileptiform activity spread across most of the slice. Largest amplitudes were located in layer II/III. Epileptiform activity was characterized by long-lasting potentials with slow rising phases and a low velocity of spread in the horizontal direction (0.044 m/s). This spatio-temporal pattern of epileptiform activity in human slices was similar to that found previously in neocortical slices from guinea pigs with bicuculline. In four of nine human slices investigated under control bath conditions (in non-epileptogenic medium), the spatio-temporal activity patterns were similar to those of guinea pigs in non-epileptogenic medium. In the remaining five human slices, however, the spread in the horizontal direction was significantly larger (4188 mm) in non-epileptogenic medium than that found in slices from guinea pigs (2171 mm). Activity in human slices showing such ‘wide spread’ in control bath conditions occasionally had characteristic features of epileptiform activity. Further work will have to clarify whether these epileptiform features reflect intrinsic epileptiform properties in human tissue slices. © 1998 Elsevier Science B.V. All rights reserved. Keywords: Spatio-temporal distribution; Epileptiform activity; Human temporal neocortex; Bicuculline; Electrical stimulation
* Corresponding author. E-mail: [email protected] 0920-1211/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII S0920-1211(98)00054-0
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
1. Introduction
2. Methods
A characteristic feature of epileptogenesis is the spread of activity across a large area of the brain. The pattern of spread has been investigated previously in different model epilepsies both in vivo (e.g. Petsche et al., 1974; London et al., 1989) and in the in vitro slice preparation (e.g. Chervin et al., 1988; Chagnac-Amitai and Connors, 1989a,b; Wadman and Gutnick, 1993). In guinea pig neocortical slices, we have shown by use of optical recordings that supragranular cortical layers are significant both for the generation and spread of epileptiform activity (Albowitz et al., 1990; Albowitz and Kuhnt, 1995). Spread was shown to be independent of the process of induction of epileptiform activity (Albowitz and Kuhnt, 1993c), and was characterized by a low velocity of horizontal (i.e. parallel to the cortical lamination) propagation (B0.04 m/s). Characterization of the spread of activity in animal experiments, however, does not necessarily reflect the situation in human epileptic brain tissue. More important than species differences in cortical structure and function is the acute nature of many animal epilepsy models. Chronic seizure disorder (‘epilepsy’) as seen in human patients is possibly accompanied by histopathological changes of the cortex, which might influence the pattern and mechanisms of spread. Therefore, we analyzed the spread of activity in slices made from human tissue resected during epilepsy surgery (Kato et al., 1973; Schwartzkroin and Prince, 1976). Electrophysiology of neurons in human slices was found to be in many aspects similar to those in neurons of other mammals (see e.g. Williamson, 1994). Epileptiform activity can readily be induced in human slices by application of bicuculline (Avoli and Olivier, 1989; Hwa et al., 1992) or by treatment with low-Mg2 + medium (Avoli et al., 1991; Straub et al., 1992). Using bicuculline to activate epileptiform activity, we monitored spread of this activity in slices from human temporal cortex and compared it to spread in guinea pig cortex under comparable experimental conditions.
2.1. Slice preparation
225
Slices were obtained from a small portion of human neocortical tissue which is normally excised for treatment of pharmacoresistant focal epilepsy originating in the temporal lobe. The patients (n= 6) had received a variety of antiepileptic drugs, which included one or more of the following: carbamazepine, phenytoin, valproate, vigabitrine, and gabapentin. Surface EEG recordings and magnetic resonance imaging (MRI) were performed in all patients included in this study, and the excised tissue was investigated for neuropathological features. In all patients, surface EEG recordings showed temporal lobe interictal spikes. In five out of six patients, MRI and pathology indicated hippocampal sclerosis. Three of six patients showed atrophy of the temporal lobe or hypometabolism in positron emission tomography, and two patients had cortical dysplasia. The experiments were approved by the local ethics committee, and informed consent was obtained from all patients. Slices from the inferior temporal gyrus (Fig. 1) were cut on a vibratome (400 mm). They were placed in a portable incubation chamber (Ko¨hling et al., 1995) with oxygenated (O2/CO2, 95%/5%, v/v) artificial cerebrospinal fluid (ACSF) at a temperature of 28°C and pH 7.4. The composition of
Fig. 1. Preparation of neocortical slices from human tissue resected during epilepsy surgery and from guinea pig neocortex.
226
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
the ACSF was (in mM): NaCl (124), KCl (5), NaH2PO4 (1.25), MgSO4 (1.3), CaCl2 (1), NaHCO3 (26), and glucose (10). Slices were allowed to recover for at least 2 h before they were stained with the voltage-sensitive dye and transferred to the recording chamber. In the recording chamber, the concentration of CaCl2 was raised to 2 mM and the temperature to 35°C. The results from 10 slices of human temporal neocortical tissue were compared to our previous results from studies of sensory neocortical slices of guinea pigs (Fig. 1, Albowitz et al., 1990; Albowitz and Kuhnt, 1993a,b,c).
2.2. Stimulation and recording techniques Details on the techniques used for stimulation, and for electrical and optical recording, have been provided elsewhere (Albowitz and Kuhnt, 1993a,b). Briefly, slices were stained with the voltage-sensitive dye RH795. Activity-dependent fluorescence signals were registered with a fast optical recording system based on a 10× 10 photodiode array, providing a spatial resolution of 94 mm/diode and a temporal resolution of 0.4 ms. Optical signals were corrected for dye bleaching and for staining and illumination irregularities. For data analysis, a baseline was calculated for each photodiode record as the average amplitude of the 250 data points (100 ms) preceding the stimulus. The first point after the stimulus that was followed by \100 data points that were above baseline was defined as onset of the stimulus-evoked response. For each photodiode record, a window of 150 ms was set from this onset point. From these window settings, onset latencies and mean window amplitudes were calculated for each individual photodiode record. Epileptiform activity was induced with bath application of 10 mM bicuculline methiodide and single- or double-pulse electrical stimulation of layer I or the white matter. Monopolar stimulation electrodes consisted of tungsten in glass (50 – 200 kV) and had tip diameters below 30 mm. The pulse width was 30 ms. Stimulation strength was twice the intensity needed to evoke detectable optical signals under control conditions.
The position of the photodiode array with respect to the slice and to the stimulation electrodes was determined photographically in situ after the recording procedure and correlated to cortical layers identified in the histological section prepared following the experiments.
3. Results Following electrical stimulation in both nonepileptogenic and epileptogenic medium, a decrease of fluorescence, which occurred with specific spatio-temporal patterns, was observed in all slices. This decrease of fluorescence corresponds linearly to membrane depolarization (Albowitz and Kuhnt, 1993a). Even though optical signals are recorded and analyzed in this study, we keep the electrophysiological terminology such as ‘potentials’ and ‘depolarization’.
3.1. Epileptiform acti6ity The spatio-temporal distribution of activity evoked by electrical stimulation of layer I is shown in Fig. 2. In this example, the distribution of activity under control bath conditions (top) and in 10 mM bicuculline (bottom) is compared for neocortical slices from human temporal lobe and from guinea pig brain. In each case, evoked epileptiform activity in bicuculline was distributed across most of the slice, with largest amplitudes in layer II/III (see also Fig. 4B and Fig. 5B). This general phenomenon was observed in all slices investigated from human (n= 10) and from guinea pig (n= 32, Albowitz et al., 1990; Albowitz and Kuhnt, 1993c) neocortical tissue. Potentials in human neocortical slices in 10 mM bicuculline recorded distal from the stimulation site were of long duration (\280 ms) and had a slow rising phase. When double-pulse stimulation was used, potentials did not show two distinguishable responses (Fig. 3A,B). The velocity of spread of activity in the horizontal direction (parallel to cortical layering) was rapid when measured close to the stimulation site, but significantly slower at more distant sites (Fig. 3B, Fig. 5C). These features were previously used to define epileptiform
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
227
Fig. 2. Distribution of activity in neocortical slices from human and guinea pig under control conditions and with 10 mM bicuculline. The activity following stimulation of layer I is shown in false color coding, with red as greatest and purple as lowest activity. The area of the slice from which records were obtained is indicated by frames in the schematic diagrams at the bottom. Activity was recorded as relative fluorescence change (dF/F) from slices stained with the fluorescence dye RH795. The results shown here reflect activity summated over a window of 150 ms. WM, white matter.
activity (Albowitz and Kuhnt, 1993c). A similar pattern was observed in guinea pig neocortical slices in 10 mM bicuculline (Albowitz and Kuhnt, 1993c). The fast velocity of spread seen close to the stimulation site was similar to that of activity under control conditions, whereas the slow spread in the periphery was a characteristic feature of epileptiform activity. The velocity of this slow spread in human slices was 0.0449 0.007 m/s (mean9 standard error of the mean, n = 9), similar to the value obtained from guinea pig neocortical slices (0.0429 0.004 m/s, n= 10). Epileptiform activity as defined above always
appeared earliest in supragranular layers before invading middle and deep cortical layers, even with the stimulation electrode in the white matter. In the example shown in Fig. 3A and Fig. 4, the velocity of spread in human slices in the vertical direction was 0.198 m/s. This spread is significantly faster than the velocity of spread of epileptiform activity in the horizontal direction, and similar to the vertical spread of activity under control (i.e. non-bicuculline) conditions. In neocortical slices from guinea pigs, spread of activity in the vertical direction was also faster (0.3919 0.030 m/s, n = 10) than spread in the
228
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
horizontal direction. However, the velocity of spread in the vertical direction in slices from human tissue was highly variable. Occasionally, jumps in onset latency of up to 50 ms were observed (e.g. Fig. 4C, right). Because of this high variability, spread in the vertical direction was not systematically investigated. Stimulation of the white matter in human neocortical slices, even when treated with bicuculline, only rarely (three out of 10 cases) evoked epileptiform activity. In those cases where it was evoked, the spatio-temporal distribution of epileptiform activity was the same as that evoked by layer I stimulation. In contrast, in guinea pig neocortical slices, epileptiform activity could readily be evoked by white matter stimulation. The spatio-temporal distribution of activity in guinea pigs was also similar following layer I and white matter stimulation. 3.2. Acti6ity under control bath conditions
Fig. 3. Traces from selected photodiodes (dots in the schematic diagrams at the right) following electrical stimulation of layer I in a human neocortical slice. Records taken under control conditions (light traces) and with bicuculline (heavy traces) are superimposed. Double-pulse stimulation of layer I (50-ms interval), time points of stimulation are indicated by small arrows. (A) Records of activity from photodiodes representing the vertical axis of the slice (layers I–V). (B) Records of activity from photodiodes representing the horizontal axis of the slices (along layer II/III). The position of the stimulation electrode is indicated by the large arrow. Distal from the stimulation site, activity (under both control and epileptogenic conditions) shows a slow rising phase, but does not show a double-pulse response.
During control conditions in human slices, stimulation of layer I evoked activity of lower amplitude and limited spread (Fig. 2). Double-pulse stimulation evoked two clearly distinguishable responses, showing paired pulse facilitation. As reported previously for guinea pig neocortical slices (Albowitz and Kuhnt, 1993a), each response consisted of two components: an early, fast rising part of short duration (possibly resembling non-synaptic activity) and a later slower and longer-lasting component. In the vertical direction, largest responses were recorded in layers I and II. Activity was restricted predominantly to upper cortical layers (Fig. 2, Fig. 4B). The velocity of spread in the vertical direction was similar to the vertical spread of activity under bicuculline conditions, and similar to that seen in guinea pig neocortical slices (Fig. 4C). In the horizontal direction, the range (i.e. distance) of spread along layer III differed among nine human slices investigated (Figs. 5 and 6). Two groups could be identified: five slices with ‘wide spread’ (4188 9 243 mm, n= 5) and four slices with ‘small spread’ (24469 115 mm, n= 4). The range of spread in the latter group did not significantly differ from that seen in guinea pigs (2171986 mm, n= 7).
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
229
Fig. 4. Mean window amplitudes (B) and onset latencies (C) along the vertical axis of a human slice close (left) and far (right) from the stimulation site in layer I. The position of the stimulation electrode, the cortical layers, and positions of photodiodes from which measurements were made are shown in the schematic drawing of the slice (A). The distribution of amplitudes and latencies during control conditions and with bicuculline is compared.
The difference in the range of spread in human slices with ‘small spread’ vs. ‘wide spread’ (1742 mm) was significant (P B0.001, unpaired t-test). In human slices bathed in normal medium, the velocity of spread in the horizontal direction was rapid at sites close to the stimulation electrode; in most cases, the entire area was activated almost simultaneously (Fig. 5C, central area). This was also observed for spread in the horizontal direction in guinea pig neocortical slices. However, in those human slices showing ‘wide spread’ the velocity of horizontal spread in the periphery was
occasionally much slower, and similar to that of epileptiform activity (Fig. 5C, peripheral area). In the example in Fig. 5C, the velocity of horizontal spread in the periphery was 0.031 m/s under control conditions and 0.035 m/s with bicuculline. Also, in human slices showing ‘wide spread’ under control conditions, recordings from the periphery occasionally showed characteristic properties of epileptiform activity: they had a slow rising phase and long duration and appeared as a single response following double-pulse stimulation (marked by arrows in Fig. 3B). However, epilepti-
230
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
form properties were not seen consistently in slices with wide spread. In temporally subsequent recordings from the same site, some records showed epileptiform properties while others did not. Thus, the activity of the slice appeared to ‘switch’ between two conditions.
4. Discussion
Fig. 5. Mean window amplitudes (B) and onset latencies (C) along layer II/III of a human slice following stimulation of layer I. The position of the stimulation electrode, as well as cortical layers and positions of photodiodes from which measurements were made, are shown in the schematic drawing of the slice (A). The distribution of amplitudes and latencies during control conditions and with bicuculline is compared.
The present results show that the spread of epileptiform activity induced by electrical stimulation in medium containing bicuculline, is similar in slices from human neocortex resected during epilepsy surgery and in slices from guinea pig neocortex. In both preparations, horizontal spread occurs across most of the slice, rapidly at sites close to the stimulating electrode but at a low velocity more distally. The amplitude of the response is largest in supragranular layers and it is at this level that the earliest onset latencies are found. Epileptiform potentials had a slow rising phase and were of long duration. Under the slice conditions, differences in structure and function of the cortex due to the specific species or the cortical area investigated (temporal cortex in humans vs. primary sensory cortex in guinea pigs) did not influence the pattern of epileptiform spread. In contrast to animal experiments, however, vertical spread of activity was not consistent in the human slice. This inconsistency may be due to the fact that, because of the strong gyration of the human brain, slices from human temporal lobe could not always be prepared exactly perpendicular to the cortical surface. As a consequence, vertically oriented neuronal connections might have been separated by the slicing procedure. Possibly for the same reason, epileptiform activity could not always be elicited by white matter stimulation. In the guinea pig neocortical slice, supragranular layers play a predominant role in the spread of epileptiform activity (Albowitz and Kuhnt, 1995). After separating supra- from infragranular layers by a horizontal cut, the supragranular portion was sufficient for the generation and spread of bicuculline-induced epileptiform activity. Also, spread was delayed when horizontal connections were separated in supragranular layers, but not if
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
Fig. 6. Range of spread along layer III under control conditions following stimulation of layer I in slices from guinea pigs and from two sets of human slices (means9 S.E.M.).
they were sectioned in infragranular layers. Since epileptiform activity frequently appeared earliest in supragranular layers regardless of the point of stimulation, supragranular structures were not only sufficient, but seemed to be preferred for generation and propagation of epileptiform activity. While lesion experiments have not been performed in the human slice preparation yet, the spatio-temporal pattern of epileptiform activity reported here also points to a significant role of supragranular layers. Epileptiform activity in the human slice always appeared in supragranular layers before infragranular structures were involved. This was also the case for the rare observations when epileptiform activity could be elicited by white matter stimulation. This predominant role of upper cortical layers is in contrast to findings by Connors (1984) and Chagnac-Amitai and Connors (1989b) who ascribe a pivotal role to intrinsically bursting neurons of layer V for the initiation, synchronization, and spread of epileptiform discharge. It remains to be seen whether the involvement of supra- versus infragranular layers in epileptogenesis is dependent on the specific epilepsy model used, e.g. the bicuculline versus the low-Mg2 + model. In contrast to the similarities of the spatio-temporal activity pattern of the bicuculline-induced model epilepsy in human and guinea pig neocortical slices, the activity pattern under control condi-
231
tions differed considerably. Here, features of epileptiform activity (wide spread, low velocity of spread, slow rising phases, and long-lasting potentials) were frequently present in the human but never in the guinea pig slice. Previously, abnormal responses from cortical neurons (Strowbridge et al., 1992), synchronously active neurons (Schwartzkroin and Haglund, 1986), and spontaneously occurring field potentials (Ko¨hling et al., 1995) have been observed in human slices under control conditions and might reflect epileptiform activity. These observations, and the epileptiform features seen in the present study, suggest that human neocortical slices obtained from tissue removed during epilepsy surgery do preserve features of its epileptogenicity in vitro. In the present investigation, the epileptic focus was not necessarily part of the tissue used for slice preparation. Thus, ‘epileptiform’ features of the activity seen in human tissue may reflect the consequences of chronic projection of epileptiform activity from the focus and may reflect intrinsic abnormalities (structural and/or functional) in related cortical tissue. Clarification of this issue will require further studies.
Acknowledgements
The authors would like to thank B. Herrenproth, S. Lausmann, E. Naß, U. Steveling, and A. Tlustochowski for excellent technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft (DFG 268/3 to B.A.).
References Albowitz, B., Kuhnt, U., 1993a. Evoked changes of membrane potential in guinea pig sensory neocortical slices: an analysis with voltage-sensitive dyes and a fast optical recording method. Exp. Brain Res. 93, 213 – 225. Albowitz, B., Kuhnt, U., 1993b. The contribution of intracortical connections to horizontal spread of activity in the neocortex as revealed by voltage sensitive dyes and a fast optical recording method. Eur. J. Neurosci. 5, 1349 – 1359. Albowitz, B., Kuhnt, U., 1993c. Spread of epileptiform potentials in the neocortical slice — recordings with voltage-sensitive-dyes. Brain Res. 631, 329 – 333.
232
B. Albowitz et al. / Epilepsy Research 32 (1998) 224–232
Albowitz, B., Kuhnt, U., 1995. Epileptiform activity in the guinea pig neocortical slice spreads preferentially along supragranular layers—Recordings with voltage-sensitive dyes. Eur. J. Neurosci. 7, 273–284. Albowitz, B., Kuhut, U., Ehrenreich, L., 1990. Optical recording of epileptiform voltage changes in the neocortical slice. Exp. Brain Res. 81, 241–256. Avoli, M., Olivier, A., 1989. Electrophysiological properties and synaptic responses in the deep layers of the human epileptogenic neocortex in vitro. J. Neurophysiol. 61, 589– 606. Avoli, M., Drapeau, C., Louvel, J., Pumain, R., Olivier, A., Villemure, J.-G., 1991. Epileptiform activity induced by low extracellular magnesium in the human cortex maintained in vitro. Ann. Neurol. 30, 499–596. Chagnac-Amitai, Y., Connors, B.W., 1989a. Horizontal spread of synchronized activity in neocortex and its control by GABA-mediated inhibition. J. Neurophysiol. 61, 747– 758. Chagnac-Amitai, Y., Connors, B.W., 1989b. Synchronized excitation and inhibition driven by intrinsically bursting neurons in neocortex. J. Neurophysiol. 62, 1149–1162. Chervin, R.D., Pierce, P.A., Connors, B.W., 1988. Periodicity and directionality in the propagation of epileptiform discharges across neocortex. J. Neurophysiol. 60, 1695–1713. Connors, B.W., 1984. Initiation of synchronized neuronal bursting in neocortex. Nature 310, 685–687. Hwa, G.G.C., Avoli, M., Olivier, A., Villemure, J.-G., 1992. Bicuculline-induced epileptogenesis in the human neocortex maintained in vitro. Exp. Brain Res. 83, 329–339. Kato, H., Ito, Z., Matsuoko, S., Sukurai, Y., 1973. Electrical activities in neurons in the sliced human cortex in vitro. Electroencephalogr. Clin. Neurophysiol. 35, 457–462. Ko¨hling, R., Lu¨cke, A., Straub, H., Speckmann, E.-J., Tux-
horn, I., Wolf, P., Pannek, H., Oppel, F., 1995. Spontaneously occurring sharp field potentials in human neocortical slice preparation. Pflu¨gers Arch. Eur. J. Physiol. 6 (Suppl.), R53. London, J.A., Cohen, L.B., Wu, J.-Y., 1989. Optical recordings of the cortical response to whisker stimulation before and after the addition of an epileptogenic agent. J. Neurosci. 9, 2182 – 2190. Petsche, H., Prohaska, O., Rappelberger, R., Vollmer, R., Kaiser, A., 1974. Cortical seizure patterns in multidimensional view: The information content of equipotential maps. Epilepsia 15, 439 – 463. Schwartzkroin, P.A., Haglund, M.M., 1986. Spontaneous rhythmic synchronous activity in epileptic human and normal monkey cortex. Epilepsia 27, 523 – 533. Schwartzkroin, P.A., Prince, D.A., 1976. Microphysiology of human cerebral cortex studied in vitro. Brain Res. 115, 497 – 500. Straub, H., Ko¨hling, R., Speckmann, E.-J., 1992. Low magnesium induced epileptiform discharges in neocortical slices (guinea pig): increased antiepileptic efficacy of organic calcium antagonist verapamil with elevation of extracellular K + concentration. Comp. Biochem. Physiol. 103C, 57 – 63. Strowbridge, B.W., Masukawa, L.M., Spencer, D.D., Shepherd, G.M., 1992. Hyperexcitability associated with localizable lesions in epileptic patients. Brain Res. 587, 158 – 163. Wadman, W.J., Gutnick, M.J., 1993. Non-uniform propagation of epileptiform discharge in brain slices of rat neocortex. Neuroscience 52, 255 – 262. Williamson, A., 1994. Electrophysiology of epileptic human neocortical and hippocampal neurons maintained in vitro. Clin. Neurosci. 2, 47 – 52.
.
.