Spatial and amplitude dynamics of neurostimulation: Insights from the acute intrahippocampal kainate seizure mouse model

Abstract

Objective: Neurostimulation is an emerging treatment for patients with drug-resistant epilepsy, which is used to suppress, prevent, and terminate seizure activity. Unfortunately, after implantation and despite best clinical practice, most patients continue to have persistent seizures even after years of empirical optimization. The objective of this study is to determine optimal spatial and amplitude properties of neurostimulation in inhibiting epileptiform activity in an acute hippocampal seizure model.

Methods: We performed high-throughput testing of high-frequency focal brain stimulation in the acute intrahippocampal kainic acid mouse model of status epilepticus. We evaluated combinations of six anatomic targets and three stimulus amplitudes.

Results: We found that the spike-suppressive effects of high-frequency neurostimulation are highly dependent on the stimulation amplitude and location, with higher amplitude stimulation being significantly more effective. Epileptiform spiking activity was significantly reduced with ipsilateral 250 μA stimulation of the CA1 and CA3 hippocampal regions with 21.5% and 22.2% reductions, respectively. In contrast, we found that spiking frequency and amplitude significantly increased with stimulation of the ventral hippocampal commissure. We further found spatial differences with broader effects from CA1 versus CA3 stimulation.

Significance: These findings demonstrate that the effects of therapeutic neurostimulation in an acute hippocampal seizure model are highly dependent on the location of stimulation and stimulus amplitude. We provide a platform to optimize the anti-seizure effects of neurostimulation, and demonstrate that an exploration of the large electrical parameter and location space can improve current modalities for treating epilepsy.

Plain language summary: In this study, we tested how electrical pulses in the brain can help control seizures in mice. We found that the electrode's placement and the stimulation amplitude had a large effect on outcomes. Some brain regions, notably nearby CA1 and CA3, responded positively with reduced seizure-like activities, while others showed increased activity. These findings emphasize that choosing the right spot for the electrode and adjusting the strength of electrical pulses are both crucial when considering neurostimulation treatments for epilepsy.

Keywords: brain stimulation; epilepsy; neuromodulation; seizures; status epilepticus.

FIGURE 1

Electrode implantation and intrahippocampal kainate‐induced spiking. (A) Surface electrode implants (axial view). Six stainless‐steel screw electrodes were implanted in the skull under isoflurane anesthesia. (B) Depth electrode implants (coronal views). KA injected into the right CA1/DG, followed by six intracerebral depth electrodes in the bilateral CA1, ipsilateral subiculum, CA3, ventral hippocampal commissure, and medial septum. Atlas images adapted from the original. Colored rectangles show locations based on histology (n = 5). (C) Surface recordings show the evolution of KA‐induced spiking in the ipsilateral‐frontal EEG. (D) Depth recordings from the ipsilateral CA3 show the evolution of KA‐induced spiking. (E) Spike detection on depth EEG: After common mode removal (top), the signal underwent band‐pass filtering z‐score estimation (middle). Spikes with z‐score > 5 were detected, and spikes within 0.2 s of the last detection were removed (bottom). (F) Effects of acute high‐frequency stimulation. Baseline: Spiking post‐KA infusion at ipsilateral CA3, with inset showing detected spikes. Stimulation: 5‐s trials of high‐frequency stimulation (100 Hz) at varying amplitudes (0–250 μA); EEG overwhelmed by artifacts. Response: Traces show spike frequency post‐stimulation: increased, unchanged, and decreased.

FIGURE 2

Spike characteristics during baseline recording. A total of 413 323 spikes were detected and analyzed across all trials. (A) Spike frequencies were compared at each location. Significant differences were found between spike frequencies at most sites (* P < 0.001); however, no differences were found between VHC, contralateral CA1, and ipsilateral subiculum. (B) Single and averaged waveform traces are shown for each depth location. The highest spike amplitudes are seen at the ipsilateral CA3 region. (C) Mean baseline spike amplitudes. Amplitudes at each site were significantly different (P < 0.001). MS, medial septum; VHC, ventral hippocampal commissure; SUB, subiculum; C, contralateral; I, ipsilateral. Error bars represent a 95% confidence interval.

FIGURE 3

Acute high‐frequency stimulation effects on epileptiform activity between different targets and amplitudes. (A) Percentage change in combined‐spike frequency, compared to no‐stimulation control. (B) Coefficient of variance of the inter‐spike intervals. (C) Pooled categorical outcome to maximal amplitude (250 μA) stimulation at each electrode location. The response was categorized as “Improved” if the spike frequency decreased by ≥25%; as “Worsened” if the spike frequency increased by ≥25%; otherwise, as “No Change.” CTRL: control/no‐stim. MS, medial septum; HCom, hippocampal commissure; VHC, ventral hippocampal commissure; SUB, subiculum; Contra, contralateral; Ipsi, ipsilateral. Error bars represent a 95% confidence interval. (*P < 0.05; **P < 0.01).

FIGURE 4

Spike‐time density before and after stimulation between different targets and amplitudes. Spikes were binned into 1‐s epochs and averaged across all trials for each combination of stimulus and location. The height of the density plots corresponds to a relative change compared to the mean baseline spike frequency. Stimulation locations are listed on the left, and stimulation amplitudes are on the bottom. The black density plot represents the combined‐spiking activity pre‐stimulation, whereas the color‐coded density plot represents the combined‐spiking activity post‐stimulation. CTRL, control/no stim; MS, medial septum; VHC, ventral hippocampal commissure; SUB, subiculum, Contra: contralateral, Ipsi: ipsilateral.

FIGURE 5

Spatial response to stimulation for each site using effective stimulation parameters. (A) Heatmap showing the change in spike frequency at each recording site when stimulating with 250 μA at different stimulation sites. The global maximum reduction in spiking activity is seen at Ipsilateral CA3 while stimulating at that same location. (B) The spike‐suppressive effect was evaluated at each recording site in response to the two stimulation parameter sets that consistently showed a significant reduction in spiking (ipsilateral CA1 and CA3, 250 μA stimulation amplitude). Results were compared to the control. Error bars represent 95% confidence intervals. (C) The effect of stimulation amplitude on spike frequency as a function of distance is shown across all six brain locations, for stimulation amplitudes of 50 μA (left), 100 μA (middle), and 250 μA (right). For the highest stimulation amplitude (250 μA), an apparent linear reduction in spiking activity was observed in ipsilateral CA1, CA3, and SUB with decreasing distance. MS, medial septum; VHC, ventral hippocampal commissure; SUB, subiculum, C, contralateral, I or Ipsi, ipsilateral.

FIGURE 6

Spike amplitude changes after maximal stimulation: comparison between different stimulation targets. (A) Post‐stimulation spike amplitudes were averaged across all recording sites and normalized to baseline. Maximal stimulation (250 μA) was applied at each location. The mean spike amplitude across all sites was increased for stimulation at VHC (**P < 0.01). CTRL: control/no stim. MS, medial septum, VHC, ventral hippocampal commissure, SUB, subiculum, C, contralateral, Ipsi, ipsilateral. Error bars represent a 95% confidence interval. (B) Example recording before and after HFS with 100 μA at the contralateral CA1. Note the increased spike frequency and amplitude post‐stimulation.