Abstracts

Rationale: We have shown that monitoring neocortical multi-neuron activity before locally-induced EEG seizures is a useful tool for studying the electrophysiological signals that precede focal seizures (Ludvig et al., 2007, Epilepsia 48 (Suppl.6) 388). These studies can lead to meaningful results only if the pre-seizure cellular signals are compared with multi-neuron recordings obtained in a spectrum of normal behaviors. Thus, the present study examined pre-seizure multi-neuron activity with an improved methodology. Methods: Four rats were implanted with an epidural cup over the right somatosensory/motor cortex, followed by the stereotaxic insertion of microelectrode arrays into layers III - V, both under the cup and outside of its rim. After recovery, behavioral, extracellular and EEG data were collected with the SciWorks system of DataWave Technologies. Data collections during normal sleep (2 min) and wakeful behaviors (2 min) were immediately followed by data collection (2 min) during and after artificial cerebrospinal fluid (ACSF; 50 microliter) delivery into the cup, as a control procedure. Next, acetylcholine (Ach; 50 or 100 mM; 50 microliter) was delivered into the same site to induce focal EEG seizures. Each solution was administered over 40 seconds with a perfusion pump. The data were analyzed with the SciWorks software and our proprietary action potential detection program. Our program recognized 500 - 5000 Hz band-passed extracellular recording segments containing abnormally large amplitude “initiator discharges” followed, within 5 msec, by a train of 2 or more large-amplitude discharges. Each of these segments was counted as a multi-neuron burst. Results: While ACSF application caused no seizure activity, epidural Ach provoked EEG seizures after an 87.4 ± 25 6 sec (mean ± S.E.M) “pre-seizure interval”. In all rats an indicator recording channel could be identified in the seizure focus where abnormally high-amplitude multi-neuron bursts (Fig. 1, asterisks) mostly appeared, in an unevenly distributed manner, in the critical “pre-seizure interval” (Fig. 1, arrowhead). The number of such bursts in this “pre-seizure interval” was 11.1 ± 3.3, significantly higher (p < 0.05; repeated measures ANOVA/Bonferroni’s test) than during sleep (1.1 ± 0.8) and ACSF exposure (2.5 ± 0.7), albeit not significantly higher than during wakeful behaviors (5.1 ± 1.6). However, when this latter value was corrected to the average shorter (87 sec) time of the pre-seizure interval, yielding 3.3 ± 1.1 bursts/wakefulness, the difference between pre-seizure and wakeful-state bursts was significant.