Figure 1.
Optogenetic stimulation of W-TChR2V4 transgenic (ChR2V-TG) rat hippocampus.
(A) Schematic drawing of the DNA fragment inserted at the site of the modified mouse Thy-1.2 expression cassette. (B) ChR2V-TG rat hippocampal organization showed that ChR2V fluorescence (green) was dense in all layers except the cell layer marked with DAPI nuclear stain (blue). (C) Schematic drawing of photostimulation to the septal hippocampus of rat brain. (D) Repetitive pulse photostimulation is schematically presented. The duty ratio was calculated as the proportion of the pulse width to the inter-pulse interval.
Figure 2.
Electrical activities of optogenetically induced seizure-like afterdischarges.
(A) Example traces of local field potentials recorded at the site of stimulation. Repetitive photostimulation at a 10 Hz frequency and a 0.05 duty ratio was applied to the hippocampus for 10 seconds by a hybrid-electrode fused with an optic fiber. The entire recording of the photostimulation and induced seizure-like afterdischarge is presented at the top. Magnified views below corresponded to the numbers. Blue dots indicate photostimuli. At the start of stimulation, only evoked potentials followed each photo pulse (1). Spontaneous activities, which were not time-locked to stimuli, emerged in addition to the evoked potentials (2) and persisted after the stimulation ended (3). The activity gradually became high in amplitude and rhythmic (4,5) then stopped spontaneously (6). Note that waveform changes were observed even under stimulation (especially in A2). (B) Raster presentation of LFPs during and immediately after stimulation demonstrates that evoked responses are gradually replaced by non-time-locked activity. (C) Simultaneous recording of optogenetically induced afterdischarges and electromyogram (EMG) of the contralateral forelimb. Clonic EMG activities appeared during stimulation, but never persisted after the LFP afterdischarge.
Figure 3.
Characteristics of optogenetically induced seizure-like afterdischarges.
(A) Average curve of frequency spectrum for the first 10 seconds of the induced afterdischarge after photostimulation. Seizure-like afterdischarges possessed frequency peaks around 10 and 25 Hz similarly after 10 Hz (blue) and 20 Hz (red) photostimulation. The duty ratio was 0.05 in boths. The power spectral density curve was an average of 10 afterdischarges recorded from 10 rats. This also indicates that the frequency characteristic was common across animals. (B) Average curve of frequency spectrum of the first 10 seconds of afterdischarges recorded in a single animal. The power spectral density curve was an average of 10 afterdischarges including 4, 5, and 1 afterdischarges after 10, 20 and 40 Hz photostimulations respectively with various duty ratios. Seizure-like afterdischarges showed frequency peaks around 10 and 25 Hz. (C) Chance of afterdischarge induction is plotted against stimulus frequency and duty ratio (n = 115 seizures, n = 10 rats). The highest chance was 1.0 (10 afterdischarges induced in 10 trials) observed with 10 and 20 Hz stimulus frequencies and a duty ratio of 0.05. Optical intensity was 19–22 mW at the tip of optic fiber and stimulus duration was 10 seconds.
Figure 4.
Afterdischarge induction was not due to an innate susceptibility of ChR2V-TG animals to seizures.
(A) AAV5-ChR2V viral vector was injected to rat hippocampus 4 weeks prior to the afterdischarge induction experiment (n = 3). (B) Fluorescent images of ChR2V protein (green) expression in fibers running through the hilus, the molecular layer of the dentate gyrus, and the striatum radiatum of CA3. (C) Raw waveform recorded in the hippocampus during afterdischarge induction. Repetitive photostimulation induced seizure-like afterdischarge similar to that observed in the experiment using ChR2V-TG rats. At least one afterdischarge was induced after photostimulation on any combinations of stimulus frequency (5, 10 ,20 Hz) and duty ratio (0.05, 0.1, 0.2). Optical intensity and stimulus duration were fixed at 19–22 mW at the tip of optic fiber and at 20 seconds, respectively. (D) Electrical stimulation was delivered to induce “classical” afterdischarge in the hippocampus. No significant difference was observed in the afterdischarge threshold between ChR2V-TG rats (n = 5) and wild-type Wistar rats (n = 5), suggesting that no inherent excitability existed in the transgenic rats. (B) Repetitive pulse photostimulation was delivered to the hippocampus and extra-hippocampal structures in ChR2V-TG rats (n = 4). The chance of afterdischarge induction was very low or zero in the amygdala, anterior thalamic nucleus and sensorimotor cortex. Stimulation parameters were set at 10 Hz with a duty ratio of 0.05. Optical intensity and stimulus duration were fixed at 19–22 mW at the tip of optic fiber and at 10 seconds, respectively. It should be noted that seizure-like afterdischarge was induced one out of 10 trials after the stimulation of the sensori-motor cortex.
Figure 5.
Optogenetic seizure-like afterdischarges activated the entire hippocampus.
(A) Fluorescent images of dentate-hilar neurons expressing ChR2V and labeled for the immediate early gene product c-Fos show robust neuronal activation produced by seizure-like afterdischarge. (B) The proportion of c-Fos positive neurons to total cells was significantly higher in the afterdischarge-induced hippocampus (n = 9) than in the control group (n = 6). No region-specific increases were observed.
Figure 6.
Causal relationships and coherences were dynamically changed along the septo-temporal axis of the hippocampus during seizure-like afterdischarges.
(A) Pulse photostimulation was delivered to the septal hippocampus, and induced afterdischarges were simultaneously recorded from multi-contact electrodes inserted along the septo-temporal axis of the temporal hippocampus (n = 3 rats). (B) Time courses of Granger causality, Granger index, and coherence are shown for both a single induced afterdischarge and a population average of afterdischarges (10 afterdischarges per rat, total 30 afterdischarges). On the population average, the thick line and shaded areas indicate the mean and 99.9% confidence interval, respectively. The average and standard deviation were calculated from the recorded LFPs of 30 trials of three rats (total 1620 LFP-pairs). The Granger causality increased in both directions but was greater in the septo-temporal direction at afterdischarge initiation. Temporo-septal causality became higher toward the end of the afterdischarge, causing transition of the Granger index to a negative value. Coherence was gradually increased toward the end of the afterdischarge. (C) State-space plots of Granger index and coherence. The population average (right) shows the mean of 30 afterdischarges. K-mean clustering (k = 3) revealed three distinct states: 1) resting state in which causality and synchrony were both low (black); 2) early phase of afterdischarge characterized by dominant septo-temporal causality and increase in coherence (red); and 3) late phase of afterdischarge characterized by reversal of causality index to the temporo-septal direction and increase in coherence (green). Transitions between phases are indicated by arrows.
Figure 7.
Causality analysis of LFPs in S-T axis of hippocampus in septal and temporal stimulations.
(A,B) Schematic of experimental set-up used for two simultaneous recording and the septal and temporal hippocampal photostimulation (n = 3, total 17 seizures). (C, D) Traces of bidirectional Granger causality (GC) scores from recorded LFPs in the stimulation to the septal (C) and temporal (D) hippocampus. (Top panels) Example traces of the GC scores. (Middle panels) Average traces of bidirectional GC scores, among three trials in one rat (septal stimulation) and seven trials of one rat (temporal stimulation). Colored regions indicate 95% confidence interval. (Bottom panels) The same as the middle panels, but the average traces were calculated using 17 trials of three rats (septal stimulation) and 18 trials of three rats (temporal stimulation). Increase of septal-to-temporal GC was observed both during the septal and temporal stimulations.
Figure 8.
State transitions during hippocampal seizure-like afterdischarges.
Three discrete states are illustrated by means of causality and coherence. Bidirectional networks along the longitudinal hippocampus work hierarchically in the genesis and termination of seizure-like afterdischarges.