Optogenetic dissection of ictal propagation in the hippocampal-entorhinal cortex structures

Abstract

Temporal lobe epilepsy (TLE) is one of the most common drug-resistant forms of epilepsy in adults and usually originates in the hippocampal formations. However, both the network mechanisms that support the seizure spread and the exact directions of ictal propagation remain largely unknown. Here we report the dissection of ictal propagation in the hippocampal-entorhinal cortex (HP-EC) structures using optogenetic methods in multiple brain regions of a kainic acid-induced model of TLE in VGAT-ChR2 transgenic mice. We perform highly temporally precise cross-area analyses of epileptic neuronal networks and find a feed-forward propagation pathway of ictal discharges from the dentate gyrus/hilus (DGH) to the medial entorhinal cortex, instead of a re-entrant loop. We also demonstrate that activating DGH GABAergic interneurons can significantly inhibit the spread of ictal seizures and largely rescue behavioural deficits in kainate-exposed animals. These findings may shed light on future therapeutic treatments of TLE.

Figure 1. Spike-triggered average of LFP (stLFP) indicates the propagating direction of ictal seizures.

(a–d) Raw spikes (upper panels) and LFPs (lower panels) recorded from representative channels in the DGH (a,c) and MEC (b,d) before (a,b) and after (c,d) kainate. (e–h) DGH spike-triggered averages of DGH LFPs (e,g: DGH–DGH stLFPs, n=5 mice) and MEC spike-triggered averages of MEC LFPs (f,h; MEC–MEC stLFPs, n=6 mice) before (e,f) and after (g,h) kainate. (i–l): Cross-area DGH spike-triggered averages of MEC LFPs (i,k; DGH–MEC stLFPs, n=5 mice) and MEC spike-triggered averages of DGH LFPs (j,l; MEC–DGH stLFPs, n=6 mice) before (i,j) and after (k,l) kainate. (m,n) Averaged peak aptitudes (m) and latencies (n) in stLFPs. Error bars represent s.d. #Represents data not available. (o) Proposed propagation direction of ictal seizures according to the stLFPs.

Figure 2. Selectively activating GABAergic interneurons inhibits multi-unit activity during ictal seizures.

(a,f) Targeting sites for optical stimulation and electrical recording in vivo. (b,d,g,i) Representative examples of multi-unit firing rates in the DGH (b,g) and MEC (d,i) at the indicated times before, during and after 60 s optical stimulation. (c,e,h,j) Averaged multi-unit firing rates in the DGH (c, n=61 from 5 mice; h, n=47 from 5 mice) and MEC (e, n=48 from 5 mice; j, n=43 from 5 mice), respectively. Light pulses (473 nm, 5 ms pulse duration at 130 Hz) were delivered into the DGH (a–e) and MEC (f–j) at time 0. The thick blue line denotes the 60-s stimulation period. Error bars represent s.e.m. Stars represent significant differences (*P<0.05, **P<0.01, ***P<0.001, paired t-test.).

Figure 3. Selectively activating GABAergic interneurons suppress ictal seizure activities.

(a,e) The targeting sites for optical stimulation and electrical recording in kainite mice. (b–d,f–h) Mean spectrograms of LFPs recorded in the DGH (b,f), MEC (c,g) and M1 (d,h), respectively (n=10 mice). Data were normalized to the maximal log₁₀(power) value across the whole 4–100 Hz frequency interval. Purple traces are raw LFP data from representative channels recorded in the DGH (b,f), MEC (c,g) and M1 (d,h), respectively. Light pulses (473 nm, 5 ms pulse duration at 130 Hz) were delivered into the DGH (a–e) and MEC (f–j) at time 0. The thick blue line denotes the initial 5 s of stimulation periods.

Figure 4. The cross-correlation of instantaneous amplitudes of field potential oscillations between DGH and MEC indicates the propagating direction of ictal seizures.

(a,b) The average distributions (a) and lags (b) of MEC–DGH cross-correlation peaks during normal state (black traces, n=3 mice), stimulation in the DGH (blue traces, n=3 mice) and stimulation in the MEC (green traces, n=3 mice), respectively. (c,d): The average distributions (c) and lags (d) of MEC–DGH cross-correlation peaks of kainate mice during normal state (black traces, n=5 mice), ictal state (red traces, n=5 mice) and stimulation in the DGH (blue traces, n=5 mice) and MEC (green traces, n=5 mice). Triangles represent the maximum points. Error bars represent s.e.m. Stars represent significant differences (*P<0.05, **P<0.01, ***P<0.001, one-sample t-test).

Figure 5. Selectively activating DGH GABAergic interneurons decreases beta–gamma-band LFP coherence.

(a,b,e,f) LFP coherence spectrum (a,e) and mean coherence (b,f) between the DGH and MEC before kainate administration (n=10 mice). (c,d,g,h) LFP coherence spectrum (c,g) and mean coherence (d,h) between the DGH and MEC after kainate administration (n=10 mice). Light pulses (473 nm, 5 ms pulse duration at 130 Hz) were delivered into the DGH (a–d) and MEC (e–h) at time 0.Insets show the targeting sites for optical stimulation and electrical recording. Thick blue lines denote the 60-s stimulation periods. Error bars represent s.e.m. Stars represent significant differences (P<0.01, paired t-test).

Figure 6. Cyclic optical activation of DGH GABAergic interneurons has a long-lasting inhibitory effect on seizures.

(a,b) Representative examples of spectrograms of LFPs in DGH of non-stimulated (a) and optically stimulated (b) mice after KA injection. Data were normalized to the maximal log₁₀(power) value across the whole 0–80 Hz frequency interval. Each blue arrow represents an optical stimulation cycle (473 nm, 5 ms pulse duration at 130 Hz, 1 min on, 5 min off); a total of 16 cycles were delivered within a 90-min period. Power values were normalized to the total power in the pre-KA period. (c) Power quantification of LFPs in the DGH of non-stimulated (n=7 mice) and optically stimulated (n=8 mice) group after kainate. (d) The seizure numbers for Racine grade IV–V behaviours calculated in each 30-min segment during a 120-min period after KA injection. (e) Normalized seizure numbers in the stimulated (n=7 mice) and non-stimulated (n=7 mice) groups. The shaded area shows the 90 min cyclic stimulation period. Error bars represent s.e.m. Stars represent significant differences (*P<0.05, **P<0.01, ***P<0.001, paired t-test).

Figure 7. Cyclic optical activation of DGH GABAergic interneurons decreased the expression of cFos in the hippocampal–entorhinal cortex (HP–EC) region.

(a) cFos expressions in non-stimulated mice 2 h after KA injection (‘After KA, No stim.' group; n=20 slices from 4 mice). (b) cFos expression in the stimulated mice 2 h after KA injection (‘After KA, Stim.' group; n=20 slices from 4 mice). (c) cFos expression in the stimulated group without KA injection (‘No KA, Stim.' group; n=20 slices from 4 mice). Light pulses (473 nm, 5 ms pulse duration at 130 Hz, 1 min on, 5 min off) were delivered to the DGH during a 90-min period. Colabeling of cFos (red) and DAPI (blue) in the HP–EC regions (left), DGH (middle) and MEC (right) is shown. (d,e) Averaged cFos/DAPI densities in the DGH (d) and MEC (e). Error bars represent s.e.m. Stars represent significant differences (*P<0.05, **P<0.01, ***P<0.001, paired t-test).