PMC

Both EEG- and action potential recordings from zebrafish larvae can be obtained simultaneously and correlated with each other. These recordings are taken from one channel following PTZ- administration. a: Data are filtered to show an EEG (bandpass at 1 to 25 Hz). b: Corresponding multi-unit activity (high pass at 100 Hz).

Spike rate plotted as a function of time (one channel shown). Recordings are made before and after PTZ addition. The addition of 0.4 mM TTX at 3900 s abolished firing. These experiments demonstrate a high likelihood that recorded signals arise from neurons firing action potentials (in contrast to myogenic potentials arising from tail-movements).

a: Baseline activity before PTZ addition. b: A typical seizure starting with a prolonged action potential burst 1–2 minutes after PTZ application. c: Seizures continue as short paroxysmal action potential bursts. PTZ addition leads to a seizure firing pattern. Markings denote sustained (> 2 sec; b) versus short (< 500 ms; c) bursting.

Top panels in (a) show representative single cell recordings before and after KCl addition, top panels in (b) before and after PTZ addition. Bottom panels show averages of baseline and KCl recordings (a) and of baseline and PTZ recordings (b). Dashed lines indicate the width of the averaged action potential at half maximum of the amplitude and they approximate 1.3 ms in both recordings. Action potentials are stable over time as indicated by the unchanged shape of the traces.

a: Representative firing fraction, a measure of percent firing channels per time, shows a significant increase in the number of channels that fire simultaneously after addition of KCl. b: The multi-channel activity is a representation of 1 sec frames displaying asynchronous firing between channels at baseline which is synchronized and highly coherent across multiple channels in the recording grid after adding KCl to the bath.

a: Summary statistics of recordings from all active channels show a significant increase (** p < 0.01) in spike rate across all experiments following KCl (p = 0.001) or PTZ addition (p = 0.001) to the bath in contrast to the sham condition (p = 0.87). b: Summary statistics of recordings from all active channels show a significant increase (* p < 0.05, *** p < 0.001) of bursting across all experiments following KCl (p = 0.0006) and PTZ addition (p = 0.02) to the bath in contrast to the sham-condition (p = 0.53). The addition of the chemical convulsants KCl and PTZ leads to an increase in firing and bursting of action potentials.

Spike rate is plotted as a function of time (bin size: 1s)–one channel per condition is shown. No significant change in burst or firing rate follows a sham convulsant application (a). Epileptic activity is detected as increased action potential firing rate and bursting after the addition of 15 mM KCl (b) or 15 mM PTZ (c). Spike rate increases with drug addition.

a: Spike rate plotted as a function of time (one channel shown). Recordings are made from a paralyzed larva (using 60 μM α-bungarotoxin) before and after PTZ addition. b: Spikes remained stable at the beginning of the PTZ addition and 3600 s later similar to recordings in larvae that were not paralyzed. c: Firing rate (p = 0.04, n = 5) and burst count (p = 0.02, n = 5) of recordings shown in (a) increase significantly (* p < 0.05) with PTZ addition in paralyzed larvae just as in unparalyzed animals.

a: Electrode chamber used for recordings (scale: 10 mm). b: Nylon mesh and slice anchor used to hold larva in place (scale: 5 mm). c: Placement of larva onto its dorsal side into the chamber with a drop of water (scale: 300 μm). d: Securing larva with nylon mesh (scale: 300 μm); microelectrode array is highlighted with red dots for enhanced visibility in this photo and the next. e: Larva positioned with head onto the microelectrode array (scale 300 μm). f: Timeline for experiments—each starting with a 15 min acclimation period, a 30 min control recording, and a 30 min recording time after drug application.

a: Area under the curve (AUC), a numerical integration, of the total firing fraction between baseline and post-drug periods shows significant increase (** p < 0.01, *** p < 0.001) after KCl (p = 0.0009) and PTZ (p = 0.002) addition relative to baseline. Note the similarity in the firing fraction across time when no drug (sham) was added to the bath (p = 0.86). b: Area under the curve of spatial coherence, a measure derived by comparing the correlation of coefficient of a central pixel with respect to its nearest-matching neighbor along time, also did not differ between baseline and follow-up in the sham condition (p = 0.19). Notably, more channels fired simultaneously indicating a significant increase (* p < 0.05, ** p < 0.01) in spatial coherence after KCl (p = 0.002) and PTZ (p = 0.02) administration.