Imaging of individual hippocampal seizures and the long-term effects of recurrent seizures.

Abstract

This invention confirms that ventral hippocampal ignition leads to functional reorganization of the ventral hippocampal excitatory circuits. Most notably, this results in increased connectivity with the medial prefrontal cortex, increased activation on fMRI, and increased electrophysiological activation amplitude. Evidence suggests increased anxiety after ignition. This invention provides a method for imaging a single epileptic seizure using simultaneous LFP-fMRI. Imaging the spatiotemporal dynamics of an individual epileptic seizure can characterize the propagation patterns of focal and secondary generalized seizures, thereby providing targeted interventions.

Description

Technical Field

[0001] Cross-reference to related applications

[0002] This patent application claims priority to U.S. Provisional Patent Application No. 62 / 945,012, filed on December 6, 2019, the entire contents of which are incorporated herein by reference for all purposes. Background Technology

[0003] When seizure activity intrudes into key circuits involving consciousness or cardiopulmonary regulation, the risk of accident or death increases. Little is known about the evolution and propagation of such seizures, but there is evidence that both cortical and subcortical circuits are involved. Standard methods for recording seizures (e.g., electrophysiology and optical imaging) have limited spatial coverage and cannot capture the evolution of whole-brain activity associated with certain events. For example, focal seizures that progress to bilateral tonic-clonic (FBTC) seizures (formerly known as secondary generalized seizures) are difficult to treat and significantly impact patient quality of life and safety. fMRI can provide whole-brain information and has been used to visualize both focal and generalized seizures in humans and animals, but its use for visualizing FBTC seizures is challenging due to motion artifacts caused by associated motor activity.

[0004] The relationship between seizures, ignition, and subthreshold activity has long been difficult to understand, yet it is crucial for the diagnosis and treatment of epilepsy. This article presents a novel approach that, for the first time, defines the specific relationship between seizures, ignition, and subthreshold activity. Summary of the Invention

[0005] This invention provides methods and models for analyzing events associated with epileptic seizures in the brain (e.g., events simulating epileptic seizures); including assessing the effects of functional and electrical neural circuit changes. The methods and models may include, for example, one or more of the following: single seizure analysis, subthreshold stimulus triggering, focal progression to bilateral tonic-clonic (FBTC) seizure analysis, excitatory ventral hippocampal (VH) network analysis, wandering seizure core analysis, etc.

[0006] In some embodiments, methods and models are provided for identifying and locating the core of a wandering seizure in an individual. In some embodiments, methods and models are provided for therapeutic agent design to treat seizures (including but not limited to FBTC seizures). In some embodiments, the identification of the wandering seizure core is used to identify the seizure onset zone. Analytical methods may include, but are not limited to, determining electrophysiology, such as local field potentials (LFP) and functional magnetic resonance imaging (fMRI). Models may include, but are not limited to, optogenetic models, and the use of optogenetic stimulation for ignition and seizure induction. Other stimulation methods may also be used, such as electrical, magnetic, pharmacological, etc. Stimulation methods that can be used to induce a single seizure in regions such as the hippocampus are preferred. Stimulation methods include subthreshold activity.

[0007] The data presented in this article illustrate the use of ignition and seizure-inducing models that can be analyzed via simultaneous electrophysiology and functional MRI. To image a single seizure in an animal model using simultaneous LFP-fMRI, the animal can be sedated and given a short-acting neuromuscular blocking agent to prevent movement during seizure imaging. Exemplary agents for this purpose include, but are not limited to, dexmedetomidine sedatives and vecuronium bromide. Detailed analysis can be performed by imaging individual seizure events associated with the seizure. For example, the core of slow-moving activity in the hippocampus has been shown to provide novel mechanisms for epilepsy propagation and generalization. Models include ignited animal brains, such as living organisms, which can be mammals (e.g., rodents like rats, mice, etc.), non-human primates, etc.

[0008] In some embodiments, an optogenetic ignition model of seizures is provided, wherein electrogrammatic seizures are induced in an animal model by cell-type-specific optogenetic stimulation, and the animal then provides reliable FBTC seizure induction over a prolonged period, wherein the prolonged period can be up to 2 weeks, up to 3 weeks, up to 4 weeks, up to 2 months, up to 3 months, or longer. The photoactivating peptide used for stimulation may be, for example, channel rhodopsin, including but not limited to CHR2. The photoactivating peptide may be operatively linked to a promoter expressed in excitatory hippocampal neurons. The stimulation pattern comprises a series of short, mild stimuli (e.g., about 10 Hz) below the threshold to trigger seizures and evaluate potential functional circuit changes. Long, intense stimuli (about 40 Hz) can be used to evaluate seizure circuit dynamics. Simultaneous electrophysiology and fMRI can be used to determine the ignition of excitatory neurons in the ventral hippocampus and the effect of seizure induction. Imaging of whole-brain network dynamics of a single induced seizure demonstrates the propagation of focal and FBTC seizures.

[0009] The model provided in this paper can be used to design and test therapeutic interventions, such as surgical and pharmacological interventions, to determine their impact on seizure induction and spread. The model can also be used to design medications to treat epilepsy comorbidities, such as the maximal activity changes occurring in the medial prefrontal cortex (mePFC), indicating an enhanced excitatory relationship between the ventral hippocampus (vHip) and the mePFC. Therapies designed to treat vHip-mePFC circuit dysfunction can reduce epilepsy-related comorbidities, including anxiety and cognitive deficits. In some embodiments, these variables of seizure spread are used to guide surgical targeting and treatment development for epilepsy. The findings in this paper include a slowly migratory core exhibiting high-amplitude activity accompanying seizures and frequently observed prior to seizures.

[0010] In some embodiments, improved surgical methods for epilepsy are provided. As is known in the art, such methods require reliable localization of the seizure onset zone (SOZ). This document demonstrates that the wandering hippocampal core provides a fundamental mechanism for seizure propagation, which can influence SOZ localization. The SOZ may be inactive throughout the seizure; the regional distribution of seizure activity can change rapidly; and the most active area is not typically the SOZ. Standard clinical methods for SOZ identification include, for example, single-photon emission computed tomography (SPECT), electrophysiology, etc. Improved methods for detecting the SOZ reflect the detection of the wandering core to improve SOZ localization. Attached Figure Description

[0011] Figure 1(A) Optogenetic ventral hippocampal ignition was targeted at CamKII cells in the ventral hippocampus and implanted with photoelectrodes for stimulation and electrophysiology. Electrodes were implanted in the ipsilateral medial prefrontal cortex for electrophysiological examination. Right image: Confocal image of ChR2-eYFP expression localized in the ventral hippocampus. (B) For optogenetic ignition, rats were stimulated every other day, up to 12 times a day, or until a Racine grade 5 seizure was observed. Stimulation continued until the ignition criterion was met: a grade 5 seizure was observed within the first three stimuli of the day. (C) 83% (10 / 12) of the rats were ignited on day 7, and all rats were ignited by day 11. (D) Behavioral scores of each ignited rat at each stimulation indicated successful ignition. (E) Total number of stimuli received by each rat to achieve ignition. (F) No significant change in hippocampal volume was observed in either the unignited or ignited animals. Left panel: Changes in ipsilateral hippocampal volume, t-test (t = -0.385, p = 0.703); Right panel: Changes in contralateral hippocampal volume, t-test (t = -1.361, p = 0.187). (G) At 3 and 12 weeks post-ignition, rat subgroups were tested to determine whether the ignition phenotype was retained. Rats were subjected to three stimuli. All ignited animals exhibited Racine grade 5 seizures, while control rats did not exhibit any seizure-related Racine behaviors (n = 5 per group), indicating that ignition caused persistent changes. All data are expressed as mean ± sem. A p < 0.05 threshold was used to determine statistical significance.

[0012] Figure 2Hippocampal ignition leads to changes in whole-brain hippocampal connectivity and increased anxiety. (A) CamKII cells in the ventral hippocampus were targeted for LFP recording by electrode stimulation in the ipsilateral ventral hippocampus (iVHip) and ipsilateral medial prefrontal cortex (iMePFC). 27 slices of the whole brain were imaged by fMRI. (B) Simultaneous LFP-fMRI was used to assess 10Hz VHip connectivity before and after ignition and in unignited control animals (n=12 per group). For each scan, a 5-second stimulus was given at 10Hz, once per minute, for 6 cycles. (C) Statistical t-plots of cage-matched, age-matched control animals (left panel) and ignited animals (right panel), with a t-threshold corresponding to p<0.001. Activity was observed in control animals primarily limited to iVHip, iDHip, iMePFC, iAmyg, and iSept, while activity in ignited rats included these areas and extended to other areas. (D) Whole-brain activation levels in unignited and ignited rats. Regional activation levels in individual animals were quantified and expressed as a percentage of activated regions. Volumes were ordered from the largest to the smallest average volume in ignited rats. Seven regions showed significant differences between control and ignited animals (t-test with adjusted p-values, multiple comparisons using Bonferroni-Holm correction). The largest effect was observed in iMePFC, followed by cMePFC, iFrAssC, iTeAssC, iOrFrC, iInsC, and finally iStria. (E) Regional CBV-fMRI peak amplitude response to 10 Hz stimulation in unignited and ignited animals. Five of the largest active regions (percentage of active ROI) in unignited animals were selected to determine the effect of ignition on the unignited network, and CBV amplitudes were inverted for interpretation. Significant increases in amplitude were observed in iMePFC and iAmyg after ignition. Each gray circle represents data from a single animal. (F) Simultaneous acquisition of LFP responses from iVHip and iMePFC in a single rat before (left) and after (right) ignition. Blue bars represent 10Hz optogenetic stimulation. An increase in iMePFC LFP response was observed after ignition. Gradient artifact correction and bandpass filtering were performed on LFP at 8–12 Hz. (G) Group analysis of LFP responses in unignited and ignited rats. Left: No difference in iVHip LFP response to 10Hz stimulation between control and ignited animals under both conditions (n=12 per group, F=0.59, p=0.81, two-way repeated measures ANOVA). Right: Increased iMePFC LFP response to 10Hz stimulation after ignition compared to control animals (n=11 per group, F=5.39, p=0.031, two-way repeated measures ANOVA).Post-hoc analysis showed a 6.4 ± 2.6-fold increase after ignition (paired t-test, t = 2.47, p = 0.033). To estimate LFP power, the ratio between LFP power during stimulation and LFP power 5 seconds before stimulation was calculated. For each animal at each time point, the median ratio of 6–18 stimulation blocks was used to assess time-dependent group effects. For each stimulation block, the normalized power of the block was calculated relative to the power 5 seconds before stimulation. Two animals were excluded from the prefrontal cortex analysis due to electrode damage (1 control animal) and artifact contamination (1 ignited animal). Data for these animals are available in the supplement. (H) Ten weeks after fMRI, a subset of animals underwent anxiety and depressive behavioral tests. (I) Sucrose preference test. Baseline results showed no preference for either of the two bottles containing water. Experimental measurements using a bottle containing sucrose solution and another containing water showed that both groups preferred the sucrose solution, and no difference was observed between the two groups (n = 7, 8, t-test, p > 0.05). (J) Forced swimming test. After ignition, no difference was observed between the groups (n = 7, 8, t-test, p > 0.05). (K) Open field test. Compared with unignited rats, ignited rats spent significantly less time in the center of the open field, indicating an increased anxiety phenotype after ignition (t-test, p = 0.046). All data are expressed as mean ± sem. A p < 0.05 threshold was used to determine statistical significance.

[0013] Figure 3Different propagation patterns of induced seizures in lit and unlit rats. (A) Seizure induction and assessment using simultaneous LFP-fMRI. Seizures were induced 90 s after baseline in unlit and lit animals (n=2–4 rats per rat and n=7 and 5 rats per rat, respectively). (B) Duration of induced seizures. Seizure duration was significantly longer in lit rats than in unlit rats (t-test, p<0.001). (C, D) Single seizure induced in unlit rats; numbers indicate equivalent time points for LFP and BOLD maps. (C) Overlap of LFP response with ventral hippocampal BOLD response. (D) Whole-brain BOLD response to optogenetic seizure induction. Left panel: Brain slices corresponding to the Paxinos rat brain atlas, showing coordinates relative to the anterior fontanelle. Blue circles indicate the site of seizure induction. Middle panel: Evolution of BOLD activity during and after seizure induction. Right panel: Voxel-level maximum intensity projection of the experiment. Activity was observed to be primarily confined to the ipsilateral hemisphere. (E, F) Single seizure induced in lit rats; numbers indicate equivalent time points for LFP and BOLD maps. (E) Overlap between LFP response and BOLD response in the ventral hippocampus. (F) Whole-brain BOLD response induced by optogenetic seizures. Left panel: Brain slices corresponding to the Paxinos rat brain atlas, showing coordinates relative to the anterior fontanelle. Blue circles indicate the site of seizure induction. Middle panel: Evolution of BOLD activity during and after seizure induction. Right panel: Voxel-level maximum intensity projection of the experiment. Activity propagation to both cortices was observed. BOLD images were normalized to baseline (defined as 60 s before stimulus onset). All images were displayed with a ±2% threshold for visualization.

[0014] Figure 4Ignited seizures gradually spread to both sides of the cortex, while unignited seizures remained localized. (A) Example of regional BOLD activity in a single optogenetic-induced seizure. Regions were automatically segmented into 44 brain regions using a general brain atlas. Blue bars indicate the optogenetic seizure induction period. (B) Number of activated regions in unignited seizures (n=20, from 7 rats) and ignited seizures (n=17, from 5 rats). Ignited rat seizures activated 15.5 ± 2.2 more regions than unignited animal seizures (p<0.0001). Gray circles indicate individual seizures. (C) Distribution of activated regions in unignited and ignited seizures. Each region in each group was normalized according to the total number of seizures (n=20 and n=17, respectively) for comparison. It was observed that in unignited seizures, activated regions were predominantly located in the ipsilateral hemisphere, while in ignited seizures, bilateral activation was more common. Black circles indicate the 80% threshold, ensuring reliable seizure onset time estimation for regional propagation analysis. (D) Regional propagation of activity in unlit and lit seizures. Regions are ordered from fastest to slowest, with white bars representing ipsilateral regions and black bars representing contralateral regions for visualization. Left panel: In unlit rat seizures, the most consistent region of activation is observed in the ipsilateral hemisphere. Right panel: A contrast with lit rat seizures where activity propagates to both hemispheres. Notably, activity propagates from the ipsilateral to the contralateral hemisphere. Red labels indicate regions activated in both groups. (E) Comparison of seizure onset time for common activated regions. The iMePFC activation rate in lit seizures was significantly faster than in unlit seizures, by 1.28 ± 0.46 s (t = 2.153, p = 0.044, corrected for multiple comparisons using the Bonferroni procedure). A p < 0.05 threshold was used to determine statistical significance.

[0015] Figure 5Slow-migrating hippocampal seizure cores are common in both unlit and lit rats and may serve as a major mechanism for seizure generalization. (A, B) Single seizure induced in unlit rats, with numbers indicating equivalent time points for LFP and BOLD maps. (A) LFP response overlaps with ventral hippocampal BOLD response. (B) Propagation of BOLD activity in the ipsilateral hippocampus. High-amplitude activity was observed to initiate at the stimulated VHip and shift the hippocampal pole to the dorsal region. By time point 8, activity was no longer detectable in the VHip electrode, but BOLD increase persisted in the DHip. (C, D) Single seizure induced in lit rats, with numbers indicating corresponding time points for LFP and BOLD maps. (C) LFP response overlaps with ventral hippocampal BOLD response. (D) Propagation of BOLD activity in the ipsilateral hippocampus. Similar to (A, B), BOLD and LFP in the vHip increased and continued to increase during seizure induction. High-amplitude activity shifts from the VHip (time points 3, 4) to the DHip pole (time points 5, 6). (E) Segmentation atlas of the ipsilateral hippocampus. The hippocampus is segmented into 4 regions, spanning 10 slices. (F) Activity propagation time from the ventral to the dorsal hippocampus. Peak-to-peak activity from the most ventral to the most dorsal segmentation region was used to calculate the time at which peak activity was observed in the ventral to dorsal hippocampus. Transmitted hippocampal activity was observed in 8 / 20 seizures in 4 / 7 unlit rats and 15 / 17 seizures in 5 / 5 lit rats. (G, H) Individual hippocampal voxel time series in unlit and lit animals. Time series are derived from the same seizure in (A, C) and segmented as in (E). Voxel time series are ordered from most ventral to most dorsal. (I, J) The core of a migratory hippocampal seizure is a single high-amplitude activity in the brain lasting more than 50 seconds before the activity propagates out of the hippocampus. Ignition of rats to induce a single epileptic seizure: (Time point 1) Epilepsy induction; (Time points 2-5) High-amplitude hippocampal activity was the only significant activity in the brain, lasting about 50 seconds; (Time point 6) Activity spread to the ipsilateral cortex; (Time points 7-8) Activity spread to the contralateral hippocampus; (Time point 9) Activity spread to the contralateral cortex.

[0016] Figure 6 Voxel-based intergroup differences were only significant after ignition (not before ignition). Statistical t-plots of ignition (right panel) and age-matched control animals in the same cage (left panel), with the t-threshold corresponding to p < 0.001.

[0017] Figure 7 The number of stimulations or the number of grade 5 seizures could not explain post-ignition activation. Voxel-based statistical t-plots of post-ignition activation versus the number of stimulations (left plot) and versus the number of grade 5 seizures (right plot), with t-thresholds corresponding to p < 0.001.

[0018] Figure 8 The medial prefrontal cortex showed an increased response amplitude to 10Hz stimulation of the ventral hippocampus after ignition. Regional time-series data of the ventral hippocampal loop.

[0019] Figure 9 Synchronized LFP recordings excluded from analysis. (AC) The two images above show LFP recordings obtained simultaneously via fMRI of the ventral hippocampus and ipsilateral medial prefrontal cortex. The two images below are magnified views of these recordings. Blue bars represent optogenetic stimulation (10 Hz, 7.5 ms pulse). (A) Burst after stimulus cancellation. The corresponding animals were excluded from electrophysiological and BOLD analyses. (B, C) Damage to the medial prefrontal cortex electrodes resulted in noisy recordings. Both animals were excluded from medial prefrontal cortex LFP analyses.

[0020] Figure 10 Similar electrophysiological features in seizures induced in awake and dexmedetomidine-sedated rats. (A, B) Ventral hippocampal LFP recordings of the same animal and seizures induced in awake and sedated states. Red arrows indicate the onset of large-amplitude spike activity; insets i-iii) are magnified views of the components of these seizures. Blue bars indicate the duration of optogenetic stimulation (40 Hz, 7.5 ms pulse). (C) Another set of electrophysiological recordings from different animals in awake and sedated states. (D) No difference in the proportion of stimuli leading to afterfire between awake and sedated states. Each line represents one animal. (E) No difference in the onset time of large-amplitude spike activity between awake and sedated states. (F) No shorter afterfire occurs in sedated states compared to awake states.

[0021] Figure 11 Regional spread in unlit and lit seizures (using 90% seizure onset time). Regions are ordered from fastest to slowest, with white bars representing ipsilateral regions and black bars representing contralateral regions for visualization. Left panel: In unlit rat seizures, the most consistent area of ​​activation is observed in the ipsilateral hemisphere. Right panel: Contrast with lit rat seizures where activity spreads to both hemispheres. Notably, activity spreads from the ipsilateral to the contralateral hemisphere.

[0022] Figure 12 : Regional spread of activity during unignited and ignited seizures (using 0% seizure frequency). Regions are sorted from fastest to slowest, with white bars representing ipsilateral regions and black bars representing contralateral regions for visualization.

[0023] Figure 13Different regional cross-correlation patterns during seizures in unlit and lit animals. Top: Mean regional cross-correlation matrix for unlit animals. Bottom: Mean regional cross-correlation matrix for lit animals. Detailed Implementation

[0024] definition

[0025] Before further describing embodiments of the invention, it should be understood that the invention is not limited to the specific embodiments described, as differences will certainly exist in actual implementation. It should also be understood that the terminology used herein is for describing specific embodiments only and is not intended to limit the inventive concept; the scope of the invention will be defined only by the appended claims.

[0026] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. Methods and materials similar to or equivalent to those described herein may also be used in the implementation or testing of embodiments of the invention.

[0027] It is important to note that, as used in this patent and the appended claims, the singular forms of “a,” “an,” and “the” include plural references unless the context clearly indicates otherwise. Thus, for example, “a compound” refers not only to a single compound but also to a combination of two or more compounds, and “substituent” refers to a single substituent as well as two or more substituents, and so on.

[0028] In describing this invention and claiming its rights, certain terms will be used according to the definitions listed below. It should be understood that the definitions provided herein are not intended to be mutually exclusive. Therefore, some chemical components may fall under the definition of more than one term.

[0029] The phrases “for example,” “for instance,” “such as,” or “including” as used herein are intended to introduce examples that further illustrate a more general subject matter. These examples are provided only to aid in understanding the invention and are not intended to be limiting in any way.

[0030] The terms “activator,” “antagonist,” “inhibitor,” “drug,” and “pharmaceutical activator” are used interchangeably in this document to refer to a chemical substance or compound that, when applied to a living organism (human or animal), induces a desired pharmacological and / or physiological effect through local and / or systemic action.

[0031] As used in this patent, the term "treatment" and the like refer to achieving the desired pharmacological and / or physiological effects. A preventive effect refers to complete or partial prevention of a disease or its symptoms, while a therapeutic effect refers to partial or complete cure of a disease and / or adverse reactions caused by a disease. As used in this patent, "treatment" encompasses any treatment of diseases in mammals, particularly humans, including: (a) preventing a subject from developing a disease or disease symptoms, wherein the subject may be susceptible to the disease or symptoms but has not yet been diagnosed (e.g., a disease that may be related to or caused by a primary disease); (b) suppressing the disease, i.e., preventing its development; and (c) alleviating the disease, even if the disease subsides.

[0032] "Therapeutic effective amount" or "effective amount" refers to the amount of a compound administered to a mammal or other subject in a manner sufficient to treat a disease, symptom, or condition. Therapeutic effective amount will vary depending on the compound, the disease and its severity, and the age, weight, etc., of the subject being treated.

[0033] As used herein, the term "unit dosage form" refers to a physically dispersed, single dosage form suitable for human and animal subjects, each unit containing a predetermined amount of the compound, calculated in an amount sufficient to produce the desired effect when mixed with a pharmaceutically acceptable diluent, carrier, or solvent. The strength of a unit dosage form depends on the specific compound used, the desired effect, and the pharmacokinetic properties of the compound in the host.

[0034] "Pharmaceutically acceptable excipients," "pharmaceutically acceptable diluents," "pharmaceutically acceptable carriers," and "pharmaceutical acceptable adjuvants" refer to excipients, diluents, carriers, and adjuvants that can be used to prepare pharmaceutical compositions that are generally safe, non-toxic, and biologically or otherwise undesirable, including excipients, diluents, carriers, and adjuvants acceptable for veterinary and human pharmaceutical uses. As used in this specification and claims, "pharmaceuticalally acceptable excipients, diluents, carriers, and adjuvants" includes one or more of such excipients, diluents, carriers, and adjuvants.

[0035] As used herein, "pharmaceutical composition" is intended to encompass compositions suitable for administration to subjects (e.g., mammals, particularly humans). Typically, "pharmaceutical compositions" are sterile and, preferably, free of contaminants that could cause unintended reactions in a subject (e.g., compounds in the pharmaceutical composition are pharmaceutical-grade compounds). Pharmaceutical compositions can be designed for administration to subjects or patients in need via a variety of different routes of administration, including oral, buccal, rectal, parenteral, intraperitoneal, intradermal, intratracheal, intramuscular, and subcutaneous administration.

[0036] The terms “individual,” “host,” “subject,” and “patient” used herein are used interchangeably and refer to animals, including but not limited to humans and non-human primates, including apes and humans; rodents, including rats and mice; cattle; horses; sheep; cats; dogs; birds, etc. “Mammal” refers to any one or more mammals, including, for example, dogs; cats; horses; cattle; sheep; rodents, etc., and primates, for example, non-human primates and humans. Non-human animal models (e.g., mammals, such as non-human primates, mice, rabbits, etc.) can be used for experimental research. Suitable animal models include, in particular, rodents, such as rats and mice.

[0037] The terms “determine,” “measure,” “evaluate,” and “determine” used in this article are used interchangeably and include both quantitative and qualitative determinations.

[0038] Local field potentials (LFPs) are potentials recorded in the extracellular space of brain tissue using microelectrodes (metallic, silicon, or glass micropipettes). Deep recording of LFPs occurs within cortical tissue (or other deep brain structures). LFP signals in the mammalian cortex reflect the activity of thousands of neurons and are commonly used to study underlying network dynamics such as sensory processing, motor planning, attention, memory, and perception. In recent decades, the importance of LFP signals has been further enhanced by the development of high-density silicon-based microelectrodes capable of simultaneously recording LFPs at thousands of locations across entire brain regions. LFPs can be used to guide neuroprosthetic devices because they are easier and more stable to record in a chronic environment compared to the spike activity of a single neuron.

[0039] Magnetic resonance imaging (MRI) is used to analyze neurophysical events. Specifically, MRI can be used to analyze functionally associated regions (anatomical neural networks) in the brain related to neurophysical events. Association patterns can represent the temporal and / or spatial correlations of neurophysical events. The target MRI technique is functional MRI (fMRI). With fMRI, temporal changes in image contrast are displayed using appropriate MR imaging scan sequences. Functional MRI (fMRI) measures signal changes in the brain caused by variations in neural activity. The brain is scanned rapidly (typically every 2-3 seconds) at low resolution. Increased neural activity can cause changes in the MR signal through variations in T*.sub.2. This mechanism is known as the blood oxygen level dependent (BOLD) effect. Increased neural activity leads to an increased demand for oxygen, and the vascular system actually overcompensates for this, resulting in an increase in the amount of oxygenated hemoglobin relative to deoxygenated hemoglobin. Since deoxygenated hemoglobin attenuates the MR signal, the vascular response leads to an increase in the signal associated with neural activity. The BOLD effect also allows for the generation of high-resolution 3D maps of the venous vascular system within neural tissue.

[0040] While the BOLD signal is the most commonly used method in neuroscience research on human subjects, the flexibility of MR imaging offers means to make the signal more sensitive to other aspects of blood supply. Alternative techniques employ arterial spin labeling (ASL) or weight the MRI signal via cerebral blood flow (CBF) and cerebral blood volume (CBV). The CBV method requires the injection of an MRI contrast agent currently undergoing human clinical trials. Because this method has been shown to be significantly more sensitive than BOLD in preclinical studies, it may expand the clinical application of fMRI. The CBF method provides more quantitative information than the BOLD signal, but its detection sensitivity is significantly reduced.

[0041] Epilepsy. Epilepsy is a brain disorder characterized by recurrent seizures over time. Types of epilepsy can include, but are not limited to, generalized epilepsy (e.g., childhood absence epilepsy, juvenile myoclonic epilepsy, grand mal seizures during wakefulness, West syndrome, Lennox-Gasto syndrome) and partial epilepsy (e.g., temporal lobe epilepsy, frontal lobe epilepsy, benign focal epilepsy in children).

[0042] Status epilepticus (SE). Status epilepticus (SE) can include, for example, convulsive status epilepticus, such as early status epilepticus, definitive status epilepticus, refractory status epilepticus, and ultra-refractory status epilepticus; non-convulsive status epilepticus, such as generalized status epilepticus and complex partial status epilepticus; generalized periodic epileptiform discharges; and periodic unilateral epileptiform discharges. Convulsive status epilepticus is characterized by the presence of a convulsive seizure state and can include early status epilepticus, definitive status epilepticus, refractory status epilepticus, and ultra-refractory status epilepticus. Early status epilepticus is treated with first-line therapy. Definitive status epilepticus is characterized by the persistence of the seizure state despite treatment with first-line therapy and second-line therapy. Refractory status epilepticus is characterized by the persistence of the seizure state despite treatment with first-line and second-line therapy and usually general anesthesia. Ultra-refractory status epilepticus is characterized by the persistence of the seizure state despite treatment with first-line therapy, second-line therapy, and general anesthesia for 24 hours or longer.

[0043] Nonconvulsive status epilepticus can include, for example, focal nonconvulsive status epilepticus, such as complex partial nonconvulsive status epilepticus, simple partial nonconvulsive status epilepticus, and mild nonconvulsive status epilepticus; and generalized nonconvulsive status epilepticus, such as delayed absence nonconvulsive status epilepticus, atypical absence nonconvulsive status epilepticus, or typical absence nonconvulsive status epilepticus.

[0044] A seizure is a physiological event or behavioral change that occurs following an event of abnormal electrical activity in the brain. The term "seizure" is often used interchangeably with "convulsion." A convulsion is a rapid and uncontrollable shaking of the body. During a convulsion, the body's muscles repeatedly contract and relax. Based on behavioral type and brain activity, seizures are classified into two main categories: generalized and partial (also known as focal or regional). Classifying the type of seizure helps doctors diagnose whether a patient has epilepsy.

[0045] A generalized seizure is caused by electrical impulses originating from the entire brain, while a partial seizure is caused by electrical impulses originating from a relatively small part of the brain (at least initially). The part of the brain from which a seizure occurs is sometimes referred to as a lesion.

[0046] There are several types of generalized seizures. The most common and most severe (and therefore the most well-known) is a generalized tonic-clonic seizure, also known as a grand mal seizure. In this type of seizure, the patient loses consciousness and usually collapses. After loss of consciousness, the patient's body stiffens (called the "tonic" phase of a seizure) for 30 to 60 seconds, followed by violent muscle twitching (the "clonic" phase) for 30 to 60 seconds, after which the patient enters a deep sleep (the "postictal" or post-ictal phase). During a grand mal seizure, injuries and accidents can occur, such as tongue biting and urinary incontinence.

[0047] Absence seizures cause a brief loss of consciousness (lasting only a few seconds) with little or no symptoms. Patients (most commonly children) typically stop what they are doing and stare blankly. These seizures occur and end suddenly and can happen several times a day. Patients are usually unaware that they are having a seizure, only knowing that "time has been lost."

[0048] Myoclonic seizures are characterized by occasional muscle twitching, usually occurring on both sides of the body. Patients sometimes describe these muscle twitches as brief electric shocks. In severe cases, these seizures may cause objects to fall or the person to involuntarily throw objects.

[0049] A clonic seizure is a repetitive, rhythmic muscle spasm that affects both sides of the body simultaneously.

[0050] The characteristic of a tonic seizure is muscle stiffness.

[0051] Atonic attacks are characterized by a sudden loss of muscle tone throughout the body, especially in the arms and legs, which often leads to falls.

[0052] Focal progression to bilateral tonic-clonic (FBTC) seizures originate in one area of ​​the brain and then spread to both sides of the brain as tonic-clonic seizures.

[0053] When assessing the severity of epileptic seizures, scales are commonly used to classify seizure-related behaviors. The Racine scale is the most widely used scale to describe these behaviors. The Racine scale consists of 5 levels, each classified as follows: Level 1: Mouth and facial clonic movements; Level 2: Level 1 + head nodding; Level 3: Level 2 + forelimb clonic movements; Level 4: Level 3 + standing; Level 5: Level 4 + repetitive standing and falling.

[0054] As used in this article, the term "ignition" or "ignition model" refers to a widely used model of seizure and epilepsy progression in which the duration and behavioral involvement of induced seizures increase after repeated induction of seizures. In such models, experimental animals are typically repeatedly stimulated with electricity or chemicals to induce seizures. Seizures following the first such stimulation are shorter in duration and have little or no behavioral impact compared to seizures induced by repeated stimulation. With each subsequent seizure, the accompanying behavior worsens, for example, progressing from freezes induced by early stimulation to convulsions induced by later stimulation. After repeated stimulation, the duration and severity of the accompanying behavior eventually reach a plateau (see, for example, Bertram, E., (2007) Epilepsy, 48(Supplement 2): 65-74).

[0055] Ignition can be achieved using various methods, including but not limited to electrical stimulation, optogenetics, and chemical treatment. When ignition is achieved using optogenetics, the photoactivatable protein is expressed in the target cells. The photoactivatable proteins that can be used include, but are not limited to, ChR2, VChR1, and C1V1. In some embodiments, the expression of the photoactivatable protein is targeted to the target neuron via a Ca2+ / calmodulin-dependent protein kinase II (CaMKII) promoter. In some embodiments, a polynucleotide encoding the photoactivatable protein is delivered to the hippocampus.

[0056] Animals can be stimulated up to 12 times a day, or until a Grade 5 motor seizure occurs, and a cycle of stimulation can be performed every other day for a maximum of 12 days. In some embodiments, the animal is stimulated less than 12 times per day, for example, 11 to 9, 9 to 7, 7 to 5, 5 to 3, or less than 3 times, until ignition is achieved. A maximum of 12 days of stimulation can be performed. In some embodiments, the animal receives stimulation for 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 day to achieve ignition. The stimulation interval can be 15 minutes. If a Grade 5 motor seizure is observed within the first three stimulations of the day, the individual can be considered ignited.

[0057] The seizure onset zone (SOG) is the cortical region where a clinical seizure occurs, as opposed to the epileptogenic zone, which is an essential cortical region for seizures. The SOG is typically located via scalp examination or invasive EEG techniques. The location of the SOG can also be determined by ictal single-photon emission computed tomography (SPECT). It is usually a portion of the stimulated area that generates spikes capable of producing afterfire. These consist of repetitive spikes that, when invading the cortex, have sufficient intensity to elicit clinical seizure symptoms. Invasive cortical surface electrodes record activity in extremely limited areas of the brain. By eliminating distance and insulation barriers, each electrode can record only the cortical area covered by that electrode. Therefore, invasive electrodes are inherently very sensitive for afterfire detection, but can only accurately define the origin of a seizure if they directly cover the SOG.

[0058] As used in this article, “neural activity” can refer to the electrical activity of neurons (e.g., changes in neuronal membrane potential) as well as indirect measurements of the electrical activity of one or more neurons. Therefore, neural activity can refer to changes in field potential, changes in intracellular ion concentration (e.g., intracellular calcium concentration), and magnetic resonance changes caused by the electrical activity of neurons, which can be measured via oxygen level-dependent (BOLD) signals such as in functional magnetic resonance imaging.

[0059] epileptic seizure model

[0060] This paper provides methods and models for analyzing brain circuits and regional relationships involved in epileptic seizures in vivo, particularly by imaging individual seizures to finely differentiate their effects. The methods of this invention can image the effects of seizures using any number of combinations of appropriate neuronal stimulation and neuronal activity measurement protocols as needed. The methods and models may include, for example, one or more of the following: individual seizure analysis, focal progression to bilateral tonic-clonic (FBTC) seizure analysis, excitatory ventral hippocampal (VH) network analysis, etc.

[0061] The models presented in this article typically use ignition animal models. Ignition refers to the plasticity induced by repeated seizures in a specific brain region after electrical stimulation, gradually increasing susceptibility to seizures. It ultimately leads to the establishment of spontaneous seizures and permanent status epilepticus. Ignition can be established through optogenetics or electrical stimulation.

[0062] To image a single seizure using simultaneous LFP-fMRI in an animal model, the animal can be sedated and given a short-acting neuromuscular blocking agent to prevent movement during seizure imaging. Exemplary agents for this purpose include, but are not limited to, dexmedetomidine sedatives and vecuronium bromide. Imaging of individual seizures has demonstrated that the slow-migrating active core in the hippocampus can provide novel mechanisms for seizure propagation and generalization; and the propagation of FBTC seizures can be imaged.

[0063] Seizures can be induced in ignited animals through the following methods: optogenetic stimulation, electrical stimulation, such as whole-brain electrical stimulation protocols, single-induction post-epileptic release; chemoconvulsants, such as pilocarpine, tetanus toxin, PTZ, erythrocyanine, fluteil, etc.; hydraulic shock injury; and high-intensity acoustic stimulation. In some embodiments, optogenetic stimulation is preferred.

[0064] In one embodiment, a combination of electrophysiological techniques (e.g., local field potential (LFP) and functional magnetic resonance imaging (fMRI) scanning different brain regions) is used to stimulate specific areas of an individual's brain to determine functional connectivity between the seizure propagation zone and other brain regions, and to image the motor activity of the seizure. Suitable analytical protocols include electrophysiology; photoinducible modulation of neural activity; electroencephalography (EEG) recording; functional imaging; and behavioral analysis. Electrophysiology may include single-electrode, multi-electrode, and / or field potential recording. As further described herein, photoinducible modulation of neural activity may include any suitable optogenetic approach. Functional imaging may include fMRI, as well as any functional imaging protocol using gene-encoded indicators (e.g., calcium indicators, voltage indicators, etc.). Behavioral analysis may include any suitable behavioral assays, such as those relating to arousal, memory (e.g., water maze tests), conditioned phenomena (e.g., fear conditioning), and sensory responses (responses to visual, somatosensory, auditory, gustatory, and / or olfactory cues).

[0065] Some protocols (e.g., fMRI) provide non-invasive, whole-brain measurements representing neural activity. Other protocols (e.g., electrophysiology) provide rapid measurements of neural activity at cellular resolution, as well as rapid control over these measurements. Still others (e.g., optogenetics) provide spatial localization and temporal control over the firing of action potentials within defined groups of neurons.

[0066] In some embodiments, methods are provided for specifically identifying and locating the core of a migratory seizure in an individual by inducing and propagating a single seizure in an ignited animal, which can be used to determine the seizure onset zone. In some embodiments, methods are provided for specifically identifying and locating an FBTC seizure in an individual by inducing and propagating a single seizure in an ignited animal.

[0067] In some embodiments, an optogenetic ignition model of seizures is provided, wherein electrocardiographic seizures are induced in an animal model by cell-type-specific optogenetic stimulation, and the animal then provides reliable FBTC seizure induction over a prolonged period, wherein the prolonged period may be up to 2 weeks, up to 3 weeks, up to 4 weeks, up to 2 months, up to 3 months, or longer. The photoactivating peptide used for stimulation may be, for example, channel rhodopsin, including but not limited to CHR2. The photoactivating peptide may be operatively linked to a promoter expressed in excitatory hippocampal neurons. The stimulation pattern comprises a series of short, mild stimuli (e.g., about 10 Hz) below the threshold to trigger seizures and evaluate potential functional circuit changes. Long, intense stimuli (about 40 Hz) may be used to evaluate seizure circuit dynamics.

[0068] In some embodiments, an optogenetic ignition model of seizures is provided, wherein electrogrammatic seizures are induced in an animal model by cell-type-specific optogenetic stimulation, and the animal then provides reliable FBTC seizure induction over a prolonged period, wherein the prolonged period can be up to 2 weeks, up to 3 weeks, up to 4 weeks, up to 2 months, up to 3 months, or longer. Stimulation patterns include a series of short, mild stimuli below the threshold (e.g., about 8–12 Hz, e.g., about 10 Hz) to trigger seizures and evaluate potential functional circuit changes. Long, intense stimuli (e.g., about 35–45 Hz, e.g., 40 Hz) can be used to evaluate seizure circuit dynamics. Simultaneous electrophysiology and fMRI can be used to determine the effect of igniting excitatory neurons in the ventral hippocampus. Imaging of whole-brain network dynamics of a single induced seizure demonstrates the propagation of focal and FBTC seizures. In some embodiments, these variables of seizure propagation are used to guide surgical targeting and therapeutic development for epilepsy. Results include a slowly migratory core exhibiting high-amplitude activity accompanying seizures and often observed prior to the seizure. The animal models can be used to design and test therapeutic interventions, such as surgery and pharmacological therapies, to determine their impact on the spread of epileptic seizures. These animal models can also be used to design drugs to treat epilepsy comorbidities, such as the maximal activity changes occurring in the medial prefrontal cortex (mePFC), indicating an enhanced excitatory relationship between the ventral hippocampus (vHip) and the mePFC. Therapies designed to treat vHip-mePFC circuit dysfunction can reduce epilepsy-related comorbidities, including anxiety and cognitive deficits.

[0069] In some embodiments, the animal model is a lit animal model. Ignition can be achieved using various methods, including but not limited to electrical stimulation, optogenetics, and chemical treatment. When ignited using optogenetics, photoactivatable proteins can be expressed in target cells. Photoactivatable proteins that can be used include, but are not limited to, ChR2, VChR1, and C1V1. In some embodiments, the photoactivatable protein targets the target neuron via a Ca2+ / calmodulin-dependent protein kinase II (CaMKII) promoter. In some embodiments, the polynucleotide encoding the photoactivatable protein is delivered to the hippocampus.

[0070] To achieve ignition, the individual may be stimulated with light at a frequency of at least 30 Hz and a pulse width of 7.5 ms. In some embodiments, the individual is stimulated at a frequency greater than 30 Hz, such as 30-35 Hz, 35-40 Hz, 40-45 Hz, 45-50 Hz, or greater than 50 Hz. The individual may be stimulated up to 12 times per day to achieve ignition. In some embodiments, the subject may be stimulated less than 12 times per day, for example, 11 to 9 times, 9 to 7 times, 7 to 5 times, 5 to 3 times, or less than 3 times, until ignition is achieved. Stimulation can be performed for up to 12 days. In some embodiments, the individual may receive stimulation for 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 day.

[0071] To study changes in functional neural circuits, individuals can be stimulated with light pulses of at least 1 Hz. In some embodiments, individuals can be stimulated with light pulses exceeding 1 Hz (e.g., 1-5, 5-10, 10-15, or 15-20 Hz).

[0072] To study changes in neural circuits during epileptic seizures, individuals can be stimulated with light pulses of at least 30 Hz. In some embodiments, individuals can be stimulated with light pulses of more than 30 Hz (e.g., 30-35, 35-40, 40-45, 45-50, or greater than 50 Hz).

[0073] In some embodiments, if the duration or severity of a seizure decreases, it is determined that the medication can provide an effective targeted intervention for the seizure. In some embodiments, the severity of the seizure is determined using the Racine scale. In some embodiments, if the severity of a seizure decreases by at least one grade according to the Racine scale, it is determined that the medication can provide an effective targeted intervention. For example, reducing the severity of the seizure from Racine grade 5 to grade 4. In some embodiments, the severity of the seizure decreases by 2, 3, or 4 grades, or all seizure symptoms disappear.

[0074] Optogenetic stimulation

[0075] As described above, animal models for detecting single epileptic seizures and the resulting electrophysiology can utilize optogenetic stimulation to achieve, for example, ignition and seizure induction, wherein the neurons involved in ignition and seizure induction operatively express photoactivated peptides.

[0076] The photostimulation used to activate the photoactivated peptide may include light pulses characterized by parameters such as frequency, pulse width, duty cycle, wavelength, and intensity. In some cases, the photostimulation includes two or more sets of different light pulses, each set characterized by a different temporal pattern. The temporal pattern can be characterized by any suitable parameters, including but not limited to frequency, period (i.e., the total duration of the photostimulation), pulse width, duty cycle, etc.

[0077] The light pulses can have any suitable frequency. In some cases, the group of light pulses contains a single light pulse that continues throughout the entire duration of light stimulation. In some cases, the frequency of the group of light pulses is 0.1 Hz or higher, for example, 0.5 Hz or higher, 1 Hz or higher, 5 Hz or higher, 10 Hz or higher, 20 Hz or higher, 30 Hz or higher, 40 Hz or higher, including 50 Hz or higher, 60 Hz or higher, 70 Hz or higher, 80 Hz or higher, 90 Hz or higher, 100 Hz or higher, and the frequency is 100,000 Hz or lower, for example, 10,000 Hz or lower, 1,000 Hz or lower, 500 Hz or lower, 400 Hz or lower, 300 Hz or lower, 200 Hz or lower, including 100 Hz or lower. In some embodiments, the frequency range of the optical pulse is 0.1 to 100,000 Hz, for example, 1 to 10,000 Hz, 1 to 1,000 Hz, including 5 to 500 Hz or 10 to 100 Hz.

[0078] For example, a series of short, mild stimuli below the threshold (e.g., about 1 Hz to about 15 Hz, about 5 Hz to about 15 Hz) can be delivered to trigger a seizure, and a stimulus of about 10 Hz can be applied to evaluate potential functional circuit changes. Long, strong stimuli (about 25 to about 50 Hz, about 35 to about 45 Hz) can be delivered; for example, a stimulus of about 40 Hz can be used to evaluate the dynamics of the seizure circuit.

[0079] In some cases, the two sets of optical pulses are characterized by having different parameter values, such as different pulse widths, for example, short or long. The optical pulses can have any suitable pulse width. In some cases, the pulse width is 0.1 ms or longer, for example 0.5 ms or longer, 1 ms or longer, 3 ms or longer, 5 ms or longer, 7.5 ms or longer, 10 ms or longer, including 15 ms or longer, 20 ms or longer, 25 ms or longer, 30 ms or longer, 35 ms or longer, 40 ms or longer, 45 ms or longer, 50 ms or longer, and 500 ms or shorter, for example 100 ms or shorter, 90 ms or shorter, 80 ms or shorter, 70 ms or shorter, 60 ms or shorter, 50 ms or shorter, 45 ms or shorter, 40 ms or shorter, 35 ms or shorter, 30 ms or shorter, 25 ms or shorter, including 20 ms or shorter. In some embodiments, the pulse width ranges from 0.1 to 500 ms, for example, 0.5 to 100 ms, 1 to 80 ms, including 1 to 60 ms, 1 to 50 ms, or 1 to 30 ms.

[0080] The average power of the light pulse (measured at the tip of the optical fiber that delivers the light pulse to the brain region) can be any suitable power. In some cases, the power is 0.1 mW or more, for example, 0.5 mW or more, 1 mW or more, 1.5 mW or more, including 2 mW or more, 2.5 mW or more, 3 mW or more, 3.5 mW or more, 4 mW or more, 4.5 mW or more, 5 mW or more, and can be 1,000 mW or less, for example, 500 mW or less, 250 mW or less, 100 mW or less, 50 mW or less, 40 mW or less, 30 mW or less, 20 mW or less, 15 mW or less, including 10 mW or less or 5 mW or less. In some embodiments, the power ranges from 0.1 to 1,000 mW, for example, 0.5 to 100 mW, 0.5 to 50 mW, 1 to 20 mW, including 1 to 10 mW or 1 to 5 mW.

[0081] The wavelength and intensity of the light pulse can vary and can depend on the activation wavelength of the photoactivated peptide, the optical transparency of the brain region, the expected brain volume to be irradiated, etc.

[0082] The volume of the brain region irradiated by the light pulse can be any suitable volume. In some cases, the irradiated volume is 0.001 mm. 3 Or larger, for example, 0.005mm 3 Or larger, 0.001mm 3 Or larger, 0.005mm 3 Or larger, 0.01mm3 Or larger, 0.05mm 3 Or larger, including 0.1mm 3 Or larger, and 100mm 3 Or smaller, for example, 50mm 3 Or smaller, 20mm 3 or smaller, 10mm 3 or smaller, 5mm 3 or smaller, 1mm 3 Or smaller, including 0.1mm 3 Or smaller. In some cases, the irradiated volume ranges from 0.001 to 100 mm. 3 For example, 0.005 to 20 mm 3 0.01 to 10 mm 3 0.01 to 5 mm 3 Including 0.05 to 1 mm 3 .

[0083] Optogenetic stimulation can be performed using any suitable method. Suitable methods are described, for example, in U.S. Patent No. 8,834,546, which is incorporated herein by reference. Neurons in suitable brain regions (whose activity will be light-modulated) can be modified using appropriate methods to express the photoactivated polypeptide. In some cases, neurons in brain regions are genetically modified to express the photoactivated polypeptide. In some cases, the neurons can be genetically modified using a viral vector, such as an adeno-associated virus vector containing a nucleic acid having a nucleotide sequence encoding the photoactivated polypeptide. The viral vector may include any suitable control elements (e.g., promoters, enhancers, recombination sites, etc.) to control the expression of the photoactivated polypeptide based on cell type, timing, the presence of inducers, etc. In some cases, cell type-specific expression of the photoactivated polypeptide can be achieved using a recombination system, such as Cre-Lox recombination, Flp-FRT recombination, etc. For cell type-specific expression of genes achieved using recombination, see the following publications: for example, Fenno et al., Nature: Methods, July 2014; 11(7): 763; and Gompf et al., Frontiers in Behavioral Neuroscience, July 2, 2015; 9: 152, which are incorporated herein by reference.

[0084] Suitable neuron-specific control sequences include, but are not limited to, Ca ++ - The α subunit of the calmodulin-dependent protein kinase II (CaMKIIa) promoter (see, for example, Mayford et al. (1996), Proceedings of the National Academy of Sciences, 93:13250) to target ventral hippocampal CAMKII neurons.

[0085] In some cases, one or more optical fibers are used to irradiate the brain region containing neurons containing photoactivated peptides. The optical fiber can be configured in any suitable manner to guide light emitted from a suitable light source (e.g., a laser or light-emitting diode (LED) light source) to the brain region. The optical fiber can be any suitable optical fiber. In some cases, the optical fiber is a multimode optical fiber. The optical fiber may include a core defining a core diameter through which light from the light source passes. The optical fiber can have any suitable core diameter. In some cases, the core diameter of the optical fiber is 10 μm or greater, for example, 20 μm or greater, 30 μm or greater, 40 μm or greater, 50 μm or greater, 60 μm or greater, including 80 μm or greater, and is 1,000 μm or less, for example, 500 μm or less, 200 μm or less, 100 μm or less, including 70 μm or less. In some embodiments, the core diameter of the optical fiber ranges from 10 to 1,000 μm, for example, 20 to 500 μm, 30 to 200 μm, including 40 to 100 μm.

[0086] The fiber optic tip implanted in the target brain region can have any suitable configuration suitable for irradiating the brain region with light stimulation delivered through the fiber optic cable. In some cases, the fiber optic cable includes an attachment device located at or near the distal end of the fiber optic cable, wherein the distal end of the fiber optic cable corresponds to the tip inserted into the subject. In some cases, the attachment device is configured to connect to the fiber optic cable and facilitate attachment of the fiber optic cable to the subject, such as to the subject's skull. Any suitable attachment device can be used. In some cases, the attachment device includes a collar, such as a metal, ceramic, or plastic collar. The collar can have any suitable size for securing and attaching the fiber optic cable.

[0087] In some embodiments, any suitable electronic components can be used to implement the method of the invention to control and / or coordinate various optical components used to irradiate the brain region. The optical components (e.g., light sources, optical fibers, lenses, objectives, mirrors, etc.) can be controlled by a controller, for example, to coordinate the irradiation of the brain region with light pulses. The controller may include a driver for the light source capable of controlling one or more parameters associated with the light pulses, such as (but not limited to) the frequency, pulse width, duty cycle, wavelength, intensity, etc. of the light pulses. The controller can communicate with components of the light source (e.g., collimators, shutters, filter wheels, moving mirrors, lenses, etc.).

[0088] Many photoactivated peptides are known in the art to be suitable for optogenetic applications, including, for example, photoactivated ion channels or photoactivated ion pumps. See, for example, Repina et al., Annual Review of Chemical and Biomolecular Engineering, June 7, 2017; 8:13-39; WO2014144409A1; US10,220,092; US10,371,776; PCT / US2011 / 028893; WO / 2013 / 093463; WO / 2017 / 210664; WO / 2017 / 015395; WO / 2019 / 092564; WO / 2017 / 100058, all of which are expressly incorporated herein by reference. The photoactivated peptides are activated by light of different wavelengths, including, for example, blue light; green light; yellow light; orange light; and red light. The photoactivated polypeptide can be fused to various sequences, such as signal peptides, endoplasmic reticulum (ER) output signals, membrane transport signals, and / or N-terminal Golgi output signals; including the addition of transport signals (ts) that enhance the transport of the protein to the cytoplasmic membrane.

[0089] The target photoactivated peptide includes, for example, step-function opsin (SFO)6 protein or stable step-function opsin (SSFO) protein, which may have specific amino acid substitutions at key positions in the retinaldehyde binding pouch of the protein. See, for example, WO 2010 / 056970, the contents of which are incorporated herein by reference in their entirety. The peptide may be a cation channel derived from Volvox (VChR1) and optionally contain one or more amino acid substitutions, such as C123A; C123S; D151A, etc. The photoactivated cation channel protein may be a C1V1 chimeric protein of the VChR1 protein derived from Volvox and the ChR1 protein derived from Chlamydomonas reinhardtii, wherein the protein comprises the amino acid sequence of VChR1 having at least the first and second transmembrane helices replaced by the first and second transmembrane helices of ChR1, optionally having an amino acid substitution at amino acid residue E122 or E162. In other embodiments, the photoactivated cation channel protein is a C1C2 chimeric protein derived from the ChR1 and ChR2 proteins of *Chlamydomonas reinhardtii*, wherein the protein is light-responsive and mediates depolarization currents in the cell when the cell is irradiated with light. In some embodiments, the depolarized photoactivated polypeptide is a redshifted variant of the depolarized photoactivated polypeptide from *Chlamydomonas reinhardtii*; referred to as "ReaChR polypeptide" or "ReaChR protein" or "ReaChR". In some embodiments, the depolarized photoactivated polypeptide is the SdChR polypeptide derived from *Scherffelia dubia*, wherein the SdChR polypeptide is capable of transporting cations across the cell membrane when the cell is irradiated with light. In some embodiments, the depolarized photoactivated polypeptide is CnChR1 derived from *Chlamydomonas noctiluca*, wherein the CnChR1 polypeptide is capable of transporting cations across the cell membrane when the cell is irradiated with light. In some embodiments, the photoactivated cation channel protein is a CsChrimson chimeric protein derived from the CsChR protein of *Chloromonas subdivisa* and the CnChR1 protein of *Chloromonas subdivisa*, wherein the N-terminus of the protein comprises amino acid residues 1-73 of CsChR, followed by amino acid residues 79-350 of CnChR1; it is photoresponsive; and when the cell is irradiated with light, it mediates a depolarization current in the cell. In some embodiments, the depolarizing photoactivated polypeptide may be, for example, ShChR1 derived from *Trichoderma pallida*, wherein the ShChR1 polypeptide is capable of transporting cations across the cell membrane when the cell is irradiated with light.

[0090] In some embodiments, the depolarizing photoactivated peptide is derived from *Chlamydomonas reinhardtii* (CHR1, particularly CHR2), wherein when the cells are irradiated with light, the peptide is capable of transporting cations across the cell membrane and mediating depolarizing currents within the cells. In some embodiments, a humanized rhodopsin CHR2 H134R mutant fused with EYFP and driven by CaMKIIa is used for optogenetic activation. The wavelength of the light used to activate the photoactivated cation channel protein derived from *Chlamydomonas reinhardtii* can be between about 460 and about 495 nm, or the wavelength can be about 480 nm. The photoactivated cation channel protein may additionally include substitutions, deletions, and / or insertions introduced into the native amino acid sequence to increase or decrease sensitivity to light, increase or decrease sensitivity to light of a specific wavelength, and / or increase or decrease the ability of the photoactivated cation channel protein to regulate the polarization state of the cytoplasmic membrane. Furthermore, the photoactivated cation channel protein may contain one or more conserved amino acid substitutions and / or one or more non-conserved amino acid substitutions. The photoactivated proton pump protein containing substitutions, deletions, and / or insertions introduced into the natural amino acid sequence appropriately retains the ability to transport cations across the cell membrane. The protein may contain a variety of amino acid substitutions, such as one or more of H134R, T159C, L132C, E123A, etc.; the protein may also contain fluorescent proteins, such as (but not limited to) yellow fluorescent protein, red fluorescent protein, green fluorescent protein, or cyan fluorescent protein.

[0091] Drug design

[0092] This invention provides a method for optimizing a therapy, wherein the optimization involves: analyzing the impact of epileptic seizures on brain regions, and based on the information, selecting appropriate candidate drugs and treatment modalities best suited to address seizure induction and spread, while minimizing unintended toxicity. The therapy is optimized by selecting treatment regimens that minimize unintended toxicity while providing effective activity.

[0093] The model provided in this paper can be used to design and test therapeutic interventions, such as surgery and pharmacological therapies, to determine their impact on seizure induction and spread. The model can also be used to design medications to treat epilepsy comorbidities, such as the maximal activity changes occurring in the medial prefrontal cortex (mePFC), indicating an enhanced excitatory relationship between the ventral hippocampus (vHip) and the mePFC. Therapies designed to treat vHip-mePFC circuit dysfunction can reduce epilepsy-related comorbidities, including anxiety and cognitive deficits. In some embodiments, these variables of seizure spread are used to guide surgical targeting and treatment development for epilepsy. The findings in this paper include a slowly migratory core exhibiting high-amplitude activity accompanying seizures and frequently observed prior to seizures.

[0094] In some embodiments, the ability of therapeutic interventions to weaken the excitatory relationship between the ventral hippocampus (vHip) and mePFC is tested. The methods disclosed herein can also be used to analyze the effects of agents on neurons and brain regions. For example, changes in excitatory relationships following exposure to one or more test compounds can be analyzed to determine the effects of said test compounds on an individual. Such analyses can be used for a variety of purposes, such as in the development of antiepileptic therapies.

[0095] Parameters are quantifiable characteristics of cells, tissues, and organisms (especially those components that can be accurately measured). For example, parameters can be the location, intensity, duration, velocity, etc., of electrophysiological discharges, and can be imaged using fMRI, LFP, etc. Readings can include a single, definitive value, or can include a mean, median, or variance. Typically, for a parameter, multiple measurements are taken to obtain a series of parameter readouts. The corresponding values ​​are expected to vary, and standard statistical methods and common statistical methods used to provide individual values ​​are employed to obtain the range of values ​​for each parameter in the test parameter set.

[0096] The target candidate drugs are bioactive agents used in drug design, encompassing many chemical categories, primarily organic molecules (which may include organometallic molecules), inorganic molecules, gene sequences, etc. Also of interest are therapeutic interventions, such as surgery, deep brain stimulation, optogenetics, etc. An important aspect of this invention is evaluating candidate therapies with preferred biological responses.

[0097] This includes pharmacologically active drugs, genetically active molecules, etc. Target compounds include chemotherapeutic agents, anti-inflammatory agents, hormones or hormone antagonists, ion channel modifiers, and neuroactive agents. Examples of pharmaceutical formulations suitable for this invention are those described in the following publication: “Pharmacological Basis of Therapeutics,” Goodman and Gilman, McGraw-Hill, New York, NY, (1996), Ninth Edition, Sections: Drugs Acting on Synaptic and Nerve Effector Connections; Drugs Acting on the Central Nervous System; Autologous Effective Substances: Pharmacological Treatment of Inflammation; Water, Salts, and Ions; etc.

[0098] Test compounds include all the types of molecules mentioned above, and may further include samples in unknown amounts. Complex mixtures of naturally occurring compounds derived from natural sources (e.g., plants) are of interest. While many samples contain compound solutions, solid samples soluble in suitable solvents can also be measured. Target samples include environmental samples, such as groundwater, seawater, mining waste, etc.; biological samples, such as lysates prepared from crop or tissue samples; manufacturing samples, such as timelines in drug development; and compound libraries prepared for analysis; etc. Target samples also include compounds for which potential therapeutic value is being evaluated, i.e., candidate drugs.

[0099] The term "sample" also includes the aforementioned fluid to which additional components have been added, such as those affecting ionic strength, pH, total protein concentration, etc. Furthermore, the sample can be processed to achieve at least partial fractionation or concentration. To minimize degradation of the compounds, the biological sample can be stored under nitrogen, cryopreserved, or a combination thereof. The volume of sample used should be sufficient for measurable detection; typically, approximately 0.1:1 to 1 ml of biological sample is adequate.

[0100] Compounds (including candidate agents) can be obtained from a variety of sources, including synthetic or natural compound libraries. For example, numerous methods are available for the random and directed synthesis of a wide range of organic compounds (including biomolecules), including the expression of random oligonucleotides and oligopeptides. Alternatively, natural compound libraries in the form of bacterial, fungal, plant, and animal extracts can be obtained or readily generated. Furthermore, naturally or synthetically produced libraries and compounds can be readily modified using conventional chemical, physical, and biochemical methods and can be used to generate combinatorial libraries. Agents are known to be capable of direct or random chemical modifications (e.g., acylation, alkylation, esterification, amidation, etc.) to generate structural analogs.

[0101] As used herein, the term "genetic factor" refers to a polynucleotide or analogue, which is tested in the screening assays of this invention by adding the genetic factor to cells. The introduction of the genetic factor alters the overall genetic composition of the cell. As used herein, the genetic factor can induce protein expression and its effects on one or more target pathways are being evaluated. The genetic factor (e.g., DNA) induces experimentally introduced changes in the cell genome, typically by integrating the sequence into a chromosome. Genetic changes can also be transient, where the exogenous sequence is not integrated but remains as an episome. RNA viruses containing the target gene and which are reverse transcribed and inserted into the host cell genome can be used. Genetic factors (peptides or polynucleotides) can also be synthesized in vitro and delivered into cells by conjugation with a portion that induces the transfer of the agent into the target cell (e.g., the 16-amino acid "Penetratin-1 peptide" from the antennae, available from Qbiogene). The genetic factor acts to increase the expression of a specific gene product in the cell and may increase and / or decrease other products in the cell.

[0102] In some cases, chemical agents with known or unknown activity are administered to animals, and their effects on seizure induction, spread, and movement are evaluated. These chemical agents can be used to activate pathways, inhibit pathways, etc., with a focus on regulating pathways other than the target pathway, and chemical agents may be more suitable than natural factors. The chemical agents are suitably added in solution or readily soluble form and can be administered to animals in various ways, such as oral, subcutaneous, and intubation methods known in the art. Preferred chemical agent formulations consist essentially of a biologically active compound and a physiologically acceptable carrier (e.g., water, saline, etc.).

[0103] In one embodiment, a combination of electrophysiological techniques (e.g., local field potentials (LFP) and functional magnetic resonance imaging (fMRI) scanning different brain regions) is used to stimulate specific areas of an individual's brain to determine the functional connectivity between the seizure propagation zone and other brain regions, and to image the movement of the seizure. The animal may be sedated, for example with dexmedetomidine; and given a short-acting neuromuscular blocking agent, such as vecuronium bromide, to prevent movement during seizure imaging with simultaneous LFP-fMRI.

[0104] Suitable analytical protocols include electrophysiology; photoinducible modulation of neural activity; electroencephalography (EEG) recording; functional imaging; and behavioral analysis. Electrophysiology may include single-electrode, multi-electrode, and / or field potential recording. As further described herein, photoinducible modulation of neural activity may include any suitable optogenetic approach. Functional imaging may include fMRI, as well as any functional imaging protocol that uses gene-encoded indicators (e.g., calcium indicators, voltage indicators, etc.). Behavioral analysis may include any suitable behavioral assays, such as those relating to arousal, memory (e.g., water maze tests), conditioning (e.g., fear conditioning), and sensory responses (responses to visual, somatosensory, auditory, gustatory, and / or olfactory cues).

[0105] The model provided in this paper can be used to design and test therapeutic interventions, such as surgery and pharmacological interventions (drug therapy), to determine their impact on seizure induction and spread. The model can also be used to design medications to treat epilepsy comorbidities, such as the maximal activity changes occurring in the medial prefrontal cortex (mePFC), indicating an enhanced excitatory relationship between the ventral hippocampus (vHip) and the mePFC. Therapies designed to treat vHip-mePFC circuit dysfunction can reduce epilepsy-related comorbidities, including anxiety and cognitive deficits. In some embodiments, these variables of seizure spread are used to guide surgical targeting and treatment development for epilepsy. The findings in this paper include a slowly migratory core exhibiting high-amplitude activity accompanying seizures and frequently observed prior to seizures.

[0106] Specific findings that can be evaluated include, for example, localization of the seizure onset zone (SOZ) using single-photon emission computed tomography (SPECT), electrophysiology, etc., where localization can detect the location of the migratory seizure core and the effect of the drug on the size, migration speed, duration, and / or location of the migratory core. High-amplitude, slowly migratory cores were found in the stimulated hippocampus, and in some cases, this was the only detectable activity prior to seizure generalization, suggesting a novel seizure generalization mechanism. The mean propagation speed of the migratory cores was 0.117 mm / s (unlit animals) and 0.107 mm / s (lit animals). The activity migrated from the iVHip to the iDHip, then spread to the ipsilateral cortex, the contralateral hippocampus, and finally to the contralateral cortex.

[0107] The induction and spread of FBTC seizures, such as the effects of medication on the size, velocity, duration, and / or location of seizures, were investigated. Results showed that ignited seizures consistently exhibited bilateral activation, while unignited seizures preferentially activated the ipsilateral hemisphere. mThal—a region considered a key node in seizure generalization—activates later than many other cortical areas. Specific analysis of mThal activation is possible.

[0108] It can be compared with known antiepileptic drugs, such as gabapentin, topiramate, lamotrigine, levetiracetam, stethamide, rufinamide, oxcarbazepine, lacosamide, perampanel, etc.

[0109] Comparison of measurement results obtained from the test and reference reagents can be accomplished using appropriate derivation schemes, AI systems, statistical comparisons, etc. The data is compared with a reference results database. A reference results database can be compiled. For each reference and test pattern, a data matrix is ​​typically generated, where each point in the data matrix corresponds to a reading of a parameter. The data for each parameter can come from repeated measurements, such as multiple individual seizures of the same type. Data points can be quantitative, semi-quantitative, or qualitative, depending on the nature of the parameter. Readings can be the mean, average, median, variance, or other statistically or mathematically derived values ​​related to the measurement. Parameter reading information can be further refined by directly comparing the corresponding reference readouts. The absolute values ​​obtained for each parameter under the same conditions will show the variability inherent in living organisms, and can also reflect individual cellular variability and inherent variability between individuals.

[0110] Classification rules are constructed based on a training dataset (i.e., a data matrix) obtained from multiple repeated experiments. Classification rules are selected to correctly identify repeated reference patterns and successfully distinguish different reference patterns. Classification rule learning algorithms can include decision tree methods, statistical methods, Naive Bayes algorithms, etc. The knowledge database will be sufficiently complex to effectively identify and classify new test drugs. Several methods for generating a sufficiently comprehensive set of classification patterns, along with sufficiently powerful mathematical / statistical methods for distinguishing them, can achieve this.

[0111] Non-pharmacological therapy design

[0112] A large number of patients still experience seizures after treatment with antiepileptic drugs, leading researchers to explore surgical and neuromodulation therapies for epilepsy. This article reviews the current status of these two treatment approaches for drug-resistant epilepsy. The effects of surgery and surgical modifications can be tested in the model described in this article to improve efficacy.

[0113] Surgical approaches include epileptirectomy, particularly for medial temporal lobe epilepsy (MTLE). However, routine localization and resection of the epileptic focus are insufficient to achieve good outcomes, and benefit may be gained from identifying the location and pathway of the core of migratory seizures. The outcome of seizures after resection is influenced by several factors, including the presence or absence of lesions on magnetic resonance imaging (MRI), the pathological matrix of the associated lesion, the extent and location of the epileptic focus, and the patient selection criteria for surgery.

[0114] Lesion transection is a concept specific to epilepsy surgery. Even if the pathogenic lesion remains in situ, complete transection of the epileptogenic cortex from the surrounding cortex and downstream midbrain is sufficient to control seizures. Hemispherectomy, prefrontal transection, and posterior transection can be used. These transection surgeries reduce surgical complications associated with extended resection.

[0115] Intracranial EEG recordings using subdural and deep electrodes can be used to identify the location of the core of a migratory seizure for surgical intervention. The use of a frameless stereotactic navigation system allows for the simultaneous and accurate implantation of both types of electrodes. While spikes and sharp waves are considered epileptogenic markers in conventional EEG, ictal DC drift and high-frequency oscillations (HFOs) are also considered epileptogenic markers in broadband EEG. HFOs are generally defined as oscillatory activity above 80 Hz. Although seizures are traditionally characterized as hypersynchronous neuronal activity, examination of ictal firing patterns of individual neurons indicates that neuronal spike activity during seizure onset and spread is highly heterogeneous and not hypersynchronous. Furthermore, as shown in this paper, a static core for generating seizures does not exist.

[0116] Non-surgical interventions include, for example, vagus nerve stimulation; deep brain stimulation (DBS) of multiple regions; closed-loop reactive stimulation; and trigeminal nerve stimulation. Vagus nerve stimulation (VNS) therapy provides long-term, intermittent stimulation of the left cervical vagus nerve (VN), generating afferent nerve impulses that stabilize the cerebral cortex and alleviate epileptic seizures.

[0117] Intracranial neurostimulation suitable for epilepsy treatment includes, for example, various targets suitable for deep brain stimulation (DBS), such as the central nucleus of the thalamus, hippocampus, subthalamic nucleus, locus coeruleus, caudate nucleus, mammillary bodies, and cerebellum. Closed-loop electrical stimulation systems hold promise for eliminating seizures by electrically stimulating the epileptic focus or other locations that respond to the detected onset of seizures.

[0118] Computer

[0119] The method of the present invention may use a computing system (e.g., a computer) to control and / or coordinate stimuli via one or more controllers and analyze data from the brain region scans. The computing unit may include any suitable components to analyze the measurement images. Therefore, the computing unit may include one or more of the following: a processor; a non-transient computer-readable storage device, such as a computer-readable medium; an input device, such as a keyboard, mouse, touchscreen, etc.; an output device, such as a display, screen, speaker, etc.; a network interface, such as a wired or wireless network interface; and so on.

[0120] Raw measurement data, such as fMRI and LFP, can be analyzed and stored on a computer-based system. As used herein, "computer-based system" refers to the hardware, software, and data storage devices used to analyze the information of this invention. The minimum hardware of the computer-based system of this invention includes a central processing unit (CPU), input devices, output devices, and a data storage device. It will be readily understood by those skilled in the art that this invention is applicable to any currently available computer-based system. The data storage device may include any manufacturer from which this information is recorded as described above, or a memory access device from which such manufacturer's information can be accessed.

[0121] Various structural formats of input and output devices can be used to input and output information in computer-based systems. This provides technicians with a similarity ranking and determines the degree of similarity contained in the test data.

[0122] The analysis can be implemented in hardware or software, or a combination of both. In one embodiment of the invention, a machine-readable storage medium is provided, comprising data storage material encoded with machine-readable data, which, when used with a machine programmed with instructions of said data, can display a comparison of any dataset with the data of the present invention. This data can be used for a variety of purposes, such as drug discovery, analysis of interactions between cellular components, etc. In some embodiments, the invention is implemented by a computer program executing on a programmable computer, said computer including a processor, a data storage system (including volatile and non-volatile memory and / or storage elements), at least one input device, and at least one output device. Program code is applied to input data to perform the functions described above and generate output information. The output information is applied to one or more output devices in a known manner. The computer can be, for example, a conventionally designed personal computer, microcomputer, or workstation.

[0123] Each program can be implemented using a high-level programming or object-oriented programming language to communicate with the computer system. However, if necessary, the program can be implemented using assembly or machine language. In any case, the language can be a compiled or interpreted language. Each such computer program can be stored on a general-purpose or special-purpose programmable computer-readable storage medium or device (e.g., ROM or disk) for configuring and operating the computer when the computer reads the storage medium or device to execute the program described herein. The system can also be considered as being implemented as a computer-readable storage medium configured with a computer program, wherein the configured storage medium enables the computer to operate in a specific and predefined manner to perform the functions described herein. Various architectural formats for input and output devices can be used to input and output information in the computer-based system of the present invention.

[0124] This document further provides methods for storing and / or transmitting sequences and other data collected using the methods disclosed herein via a computer. Any computer or computer accessory (including, but not limited to, software and storage devices) can be used to implement this invention. Sequences or other data (e.g., results of immune repertoire analysis) can be input into the computer directly or indirectly by a user. Additionally, any device used for DNA sequencing or analysis of DNA or analysis of immune repertoire data can be connected to a computer to enable data transfer to the computer and / or a computer-compatible storage device. Data can be stored on a computer or a suitable storage device (e.g., a CD). Data can also be transmitted from the computer to another computer or data collection point via methods known in the art (e.g., the Internet, terrestrial mail, airmail). Therefore, data collected by the methods described herein can be collected at any point or geographical location and transmitted to any other geographical location.

[0125] experiment

[0126] To analyze the seizure network by imaging individual hippocampal seizures before and after ignition.

[0127] In this study, we developed and utilized a novel epileptiform ignition model by refining the FBTC seizure induction technique used in epilepsy research and antiepileptic drug development. By repeatedly inducing electrocardiographic seizures using cell-type-specific optogenetic stimulation, we found that FBTC seizures occurred and could be reliably induced over several months thereafter. We then used simultaneous electrophysiology and fMRI to investigate the effects of ignition on excitatory neurons in the ventral hippocampus, the most commonly affected region in human epilepsy. We investigated the underlying hippocampal circuits altered by ignition and tested them against two common epilepsy-related comorbidities: anxiety and depression, to investigate the relationship between ignition, underlying circuits, and behavior. We directly imaged the whole-brain network dynamics of a single induced seizure to reveal the propagation of focal and first-time FBTC seizures in ignited animals. In the process, we identified key features of the seizures that are important for our understanding of seizure mechanisms, surgical targeting, and therapeutic development.

[0128] Optogenetic ignition in the ventral hippocampus. Ignition uses activity to remodel neuronal circuits. We developed a novel ignition method and investigated the effects of repeated optogenetic stimulation of the ventral hippocampus (VHip), a region commonly the origin of spontaneous seizures in human and animal models. We targeted CAMKII neurons in the rat ventral hippocampus (VHip). Figure 1 A, B). Successful ignition is characterized by the induction of a generalized motor seizure within the first three stimuli of the day (Racine grade 5).

[0129] All rats reached the ignition criterion on day 11 of stimulation, with 83% (10 out of 12 rats) reaching the ignition criterion on day 7. Figure 1 C). The average number of stimulations required is 53.6 (range: 29-117). Figure 1 D) The mean number of convulsive seizures (Racine grade 3-5) was 6.75 (range: 2-13), and the mean number of generalized seizures (Racine grade 5) was 2.9 (range: 1-9). Visualization of behavioral scores for each animal per stimulus is shown in [link to visualization]. Figure 1 E.

[0130] Compared to other common epilepsy models, ignition minimizes cell loss. To assess cell loss, we measured hippocampal volume at corresponding time points before and after ignition, and in unignited control rats, based on anatomical MRI. Consistent with other ignition models, we did not detect significant changes in hippocampal volume in either hemisphere. Figure 1F). In the ipsilateral and contralateral hippocampuses, the corresponding volume changes for the ignited and unignited animals were 0.63±0.82% and 0.14±0.96%, respectively, p=0.7 and 1.37±1.22% and -1.03±1.27%, p=0.19.

[0131] A key characteristic of the ignition model is that once an animal is ignited, the effects are clearly persistent, thus readily inducing motor seizures. To confirm whether our novel ignition procedure resulted in a persistent ignition state, we evaluated rats at 3 and 12 weeks post-ignition, as well as a subset of unignited control animals (n=5 in both groups). Motor seizures were observed only in the ignited rats and at both time points (Racine grade 5), while the unignited rats did not exhibit any marked seizure behavior at any time point (Racine grade 0). Figure 1 In summary, our findings demonstrate that this optogenetic ignition procedure shares key characteristics with conventional ignition methods and provides a new platform for our research on circuit remodeling and FBTC seizures.

[0132] Ignition resulted in extensive ventral hippocampal circuit remodeling and increased anxiety. Ignition induces local structural and functional remodeling, but it remains unclear which downstream regions are affected. We addressed this challenge by stimulating VHip CAMKII neurons at 10 Hz and evaluated whole-brain responses with and without ignition using synchronized local field potentials (LFP) and cerebral blood volume (CBV)-fMRI. To visualize the whole-brain response, we used standard generalized linear modeling methods to generate fMRI activation maps to identify voxels that were significantly modulated during stimulation. In unignited rats, stimulation-induced activity was primarily confined to the following regions in the ipsilateral hemisphere: dorsal hippocampus (DHip), VHip, septum, amygdala, and medial prefrontal cortex (MePFC). Figure 2 C, left figure). Conversely, in the lit rats, the same stimulus elicited activity beyond these areas and included areas in the contralateral hemisphere ( Figure 2 C, right figure, comparison of groups at different time points. Figure 6 In ignited rats, we tested whether individual differences in the ignition procedure led to differences in regional activity patterns. However, using the number of stimuli achieving ignition or the number of grade 5 motor seizures as regression parameters yielded no apparent voxel relationships other than noise levels. Figure 7 ).

[0133] We investigated whole-brain differences between unlit and lit rats by segmenting the brain into 44 separate regions and quantifying their activation levels. Figure 2D). After adjusting for multiple comparisons, we found seven distinct brain regions between the unlit and lit groups: ipsilateral MePFC (27.2±9.0% vs. 83.3±9.0%, p=0.011), contralateral MePFC (1.4±0.7% vs. 29.2±7.3%, p=0.033), ipsilateral frontal association cortex (iFrAssC) (0.24±0.24% vs. 27.1±5.6%, p=0.004), ipsilateral temporal association cortex (iFrAssC) ... The differences were as follows: TeAssC (0.34±0.34% vs. 26.7±4.8%, p=0.0006), ipsilateral orbitofrontal cortex (iOrFrC) (1.4±0.7% vs. 23.8±3.9%, p=0.0005), ipsilateral insular cortex (iInsC) (0.2±0.18% vs. 13.0±2.7%, p=0.0047), and ipsilateral striatum (iStria) (2.6±0.9% vs. 14.8±2.7%, p=0.011). Notably, no significant differences were observed in iVHip or iDHip (17.3±3.0% vs. 27.8±3.1%, p=0.684, 10.1±2.9% vs. 24.4±5.9%, p=0.994).

[0134] We will next investigate how ignition affects the unignited ventral hippocampal loop by comparing the regional CBV-fMRI amplitude between the two groups. Figure 2 E). We selected five ROIs with the highest activation levels in the unignited group for comparison. Figure 2 C, left figure, ROI time series (see figure below) Figure 8 After multiple comparison correction, we did not detect differences in CBV-fMRI response amplitude in iVHip (2.1±0.3% vs. 3.1±0.4%, p=0.122), iSept (1.6±0.3% vs. 2.5±0.3%, p=0.082), or iDHip (0.8±0.2% vs. 1.2±0.2%, p=0.178). However, significant differences were observed in iMePFC (1.3±0.3% vs. 4.2±0.6%, p=0.0023) and iAmyg (1.1±0.3% vs. 3.0±0.4%, p=0.0023). These data suggest that ignition enhances the response in specific downstream regions of the unignited ventral hippocampal loop.

[0135] The above results indicate that during optogenetic iVHip stimulation, optogenetic ignition in iVHip leads to a wider distribution of activity in iMePFC. Figure 2 D) The activity is more intense. Figure 2E), which was measured by fMRI. We then analyzed the synchronously acquired LFP to determine if there were corresponding LFP signal changes. Upon ignition, the response in the iMePFC significantly increased ( Figure 2 F), which is measured via LFP. We quantify this response by dividing the band power during the 5s stimulation period by the band power 5s before stimulation begins. We use this metric to compare before and after ignition, as well as between the unignited and ignited groups. We did not detect any significant differences in iVHip ( Figure 2 G, left plot, p = 0.81, n = 12 per group, two-way repeated measures ANOVA. However, in iMePFC, we found a significant interaction between group and time. Figure 2 G, right figure, p = 0.031, n = 11 per group. The reduced number of animals recorded by iMePFC is due to electrode malfunction, see [link to relevant documentation]. Figure 9 Further analysis showed that the LFP amplitude increased by 6.4 ± 2.6 times after ignition (p = 0.033, paired t-test). Combined with the fMRI response, this indicates that ignition increases the connectivity between iVHip and iMePFC.

[0136] In the lithographed animals, we found significant changes in the iMePFC response and an increase in overall brain response, but it remains unclear whether these changes will have any long-term effects on behavior. Since the ventral hippocampus regulates mood and affective behavior, we investigated whether ventral hippocampal lithographing leads to increased anxiety and depression: two of the most common comorbidities of epilepsy, with controversial mechanistic origins. Anxiety and depression have already been detected following amygdala lithographing. We investigated the long-term behavioral effects of our optogenetic lithographing on anxiety and depression by conducting behavioral tests 12 weeks after lithographing (10 weeks after the second fMRI time point) in a subset of animals. Figure 2 H). We used forced swimming and sucrose preference experiments (respectively). Figure 2 Depression was assessed using I and J tests, and anxiety was assessed using an open field test. Neither depression test showed a significant difference between unlit and lit animals. For the forced swimming test, the mean immobility score in the unlit control group was 36.3 ± 0.99, while the corresponding score in the lit rats was 34.1 ± 1.89. Figure 2 I, p = 0.303, t-test). For the sucrose preference experiment, there was no evidence of a difference between unlit and lit rats (i.e., p = 0.303, t-test). Figure 2J, no significant interaction was found between the group and the habituation / sucrose experimental period (F = 0.48, p = 0.499, two-way repeated measures ANOVA, mean estimates for unlit and lit animals were 96.2 ± 0.9% and 83.9 ± 12.8%, respectively). When sucrose water was introduced during the experiment, the total fluid consumption increased in both groups (32.3 ± 8.0 ml in the unlit rat group, p = 0.005; 24.7 ± 6.9 ml in the lit rat group, p = 0.011, paired t-test). In contrast, there was evidence of an increased anxiety phenotype in lit animals. Lighted animals spent less time in the open field center compared to unlit rats ( Figure 2 K, 12.5 ± 1.5% vs 8.0 ± 1.4% of center-time spent, p = 0.046, unlit control rats vs lit rats, t-test, n = 8 and 7). Importantly, the two groups of animals moved similar distances during the 5-minute experiment (2446 ± 153 cm vs 2203 ± 197 cm, p = 0.98). We found that ventral hippocampal ignition led to increased anxiety rather than depression, which supports the hypothesis that repetitive aberrant activity leads to underlying circuit changes that induce anxiety.

[0137] The above experiments demonstrate, as measured by open-field experiments, that optogenetic ventral hippocampal ignition leads to increased iVHip-iMePFC connectivity and increased anxiety. Notably, in previous studies, increased θ-coupling between the Vhip and MePFC was observed when these animals were placed in an anxiety-inducing region (open field) compared to placing normal rodents in a safe zone. In another study, injecting an arch into the Vhip and implanting an optical fiber into the MePFC eliminated the anxiety response in normal animals placed in an anxiety-inducing environment by inhibiting Vhip input. This is interesting because anxiety is one of the most common comorbidities of epilepsy. Our research indicates that hippocampal ignition induces changes in iVHip-iMePFC connectivity, providing a solid basis for the observation of anxiety in epilepsy.

[0138] Whole-brain imaging of single seizures in unlit and lit rats was performed using simultaneous LFP-fMRI. fMRI provides whole-brain information and has been used to visualize focal and generalized seizures in humans and animals; however, its use for visualizing FBTC seizures is challenging due to motion artifacts caused by associated motor activity. Animal fMRI typically requires sedation or anesthesia to restrict movement, but the use of sedation and anesthesia can affect seizure activity, which in turn affects associated motor activity. We have previously demonstrated that we can reliably induce seizures with our dexmedetomidine sedation protocol used for imaging, and that the electrophysiological characteristics of seizures are similar in awake and sedated states (see Electrophysiological characteristics of unlit seizures in animals in awake and sedated states for details). Figure 10 When seizures were induced in lit animals under dexmedetomidine sedation, stereotyped seizure behaviors reminiscent of those observed in awake animals were observed, indicating that the motor seizure circuitry is activated under sedation. Therefore, we developed a protocol based on dexmedetomidine sedation combined with vecuronium bromide (a short-acting neuromuscular blocking agent) to prevent movement in animals during seizure imaging with simultaneous LFP-fMRI. Figure 3 A). As expected, in lit animals, the duration of seizures estimated by iVHip LFP was longer than in unlit animals ( Figure 3 B, 66.2±6.7s and 35.4±5.17s; p=0.001).

[0139] In both lit and unlit animals, increased iVHip BOLD signaling corresponded to increased iVHip LFP amplitude associated with seizures, indicating that BOLD successfully captured focal seizure activity. Figure 3 C, E). Figure 3 An example of a single seizure with selected time points is shown, capturing the spread of activity in both unlit and lit animals. Voxel-level maximum intensity projection (MIP) during the scan provides a summary of the activated regions during the seizure. Figure 3 (D, F). These MIP activity patterns bear a striking resemblance to the corresponding patterns of focal and generalized seizures reported using endpoints such as 2-deoxyglucose.

[0140] In an example of epileptic seizures in an unlit animal ( Figure 3 C, D), the initial burst of activity in the ventral hippocampal loop is reminiscent of the 10Hz iVHip subthreshold stimulation network ( Figure 2 The activity then spreads to the ipsilateral dorsal hippocampus, followed by a negative BOLD response associated with the postictal effects.

[0141] In a single epileptic seizure example of an animal that has been ignited ( Figure 3 E, F), throughout the scanning process, the areas showing BOLD activity increased significantly, and activity can now also be detected in the cortex. Figure 3 F). Importantly, BOLD activity was observed to spread from the hippocampus to the ipsilateral cortex and then to the contralateral cortex. Figure 3 E time points 5-7) clearly demonstrate for the first time the dynamics of seizure propagation from focal to secondary generalized seizures throughout the brain.

[0142] For both unignited and ignited seizures, once activity spreads from iVHip, BOLD activity continues until the seizure activity is detected on the iVHip LFP. Figure 3 The data at time points C and D (9-10) and time points E and F (8-12) clearly demonstrate how focal electrical recordings can underestimate the duration of a seizure.

[0143] Different whole-brain propagation dynamics in unlit and lit animals. Next, we compared the regional BOLD propagation patterns of unlit and lit seizures by calculating the regional seizure onset time (n=20, from 7 unlit rats; n=17, from 5 lit rats. One unlit rat and two lit rats lost their cephalopods and could not be imaged). Using brain atlases, the mean regional response to 44 ROIs for each seizure was segmented and calculated ( Figure 4 A. Regional dynamics of a single seizure. Seizure onset time was defined as the time when BOLD activity first exceeded four standard deviations above the pre-stimulation baseline by 60 seconds, with the additional constraint that activity must remain above the threshold for at least 5 seconds within the subsequent 10 seconds. We compared the number of activated ROIs and found that seizures in the lit group involved more ROIs than in the unlit group ( Figure 4 B, 23.4 ± 2.0 regions and 38.8 ± 1.0 regions, n = 20 and 17, control and ignited, p < 0.0001).

[0144] Next, we used the improved radar chart to study the frequency of regional activation in unlit and lit seizures. Figure 4 C). This visualization of the dataset shows that ignited seizures are always bilaterally activated, while unignited seizures preferentially activate the ipsilateral hemisphere.

[0145] To calculate the average regional seizure onset time used to study the spread of activity, only regions active in at least 80% of seizures were used. Figure 4For C, n = 16–20 (unignited seizures) and n = 14–17 (ignited seizures), an 80% threshold was used to obtain a reliable estimate of seizure onset time. The 90% and 0% thresholds are shown in [reference needed]. Figure 11 , 12 In unignited seizures, at least 80% of seizures show activity in eight areas, all originating from the ipsilateral hemisphere. Figure 4 D, left image, regions ordered from fastest to slowest: iPiriC, iMePFC, iOrFrC, iVHip, iSept, iInsulC, ipsilateral entorhinal cortex (iEntC), iTeAssC. For ignited seizures, 38 regions from both hemispheres are active in at least 80% of seizures. Figure 4 (D, right figure). Furthermore, a clear transmission pattern was observed from the ipsilateral hemisphere to the contralateral hemisphere: of the 20 regions with the fastest onset time, 18 were located in the ipsilateral hemisphere. Of the remaining 18 regions, 14 were located in the contralateral hemisphere (see analysis of cross-correlation regions for each group). Figure 13 It is noteworthy that mThal—a region considered a key node in the generalization of epileptic seizures—is activated later than many other areas of the cortex, suggesting that mThal activation is a downstream activity in the generalization of epileptic seizures.

[0146] Then, we compared the regional seizure onset time of eight common active regions between the unlit and lit groups. Figure 4 E). We detected that in the iMePFC, the onset time of ignited seizures (2.8 ± 0.26 s) was faster than that of unignited seizures (4.1 ± 0.34 s, Holm's adjusted p = 0.044), while no consistent difference was observed in other regions (p > 0.3 for all other regions, unignited vs. ignited): iVHip (4.2 ± 0.43 s vs. 4.4 ± 0.76 s), iOrFrC (4.2 ± 0.51 s). The mean values ​​for iSept (3.0±0.21s), iSept (4.6±0.62s and 7.1±1.34s), iInsC (7.1±3.94s and 6.9±2.33s), iPiriC (3.4±0.32s and 3.4±0.33s), iEntC (9.8±3.06s and 9.7±5.7s), and iTeAssC (10.8±2.14s and 8.6±3.46s) are also mentioned.

[0147] High-amplitude migratory core in the hippocampus and its role in seizure generalization. During seizures, we frequently observed a slow, high-amplitude migratory core in the stimulated hippocampus; in some cases, this was the only detectable activity prior to seizure generalization, suggesting a novel seizure generalization mechanism. The migratory core was observed in the ipsilateral hippocampus of both unlit and lit groups, indicating that the core can form with or without ignition-induced circuit remodeling. This core was observed in 62% (23 / 37) of seizures in 75% (9 / 12) of rats. These numerical analyses suggest that the core was observed more frequently in lit rats (15 / 17 of 5 / 5 of lit rats) compared to unlit control rats (8 / 20 of 4 / 7 unlit rats). Figure 5 A and B demonstrate examples of the wandering core of an unlit animal. Figure 5 C and D show corresponding examples of animals that have been set on fire.

[0148] Given the core's mobility, we then divided the hippocampus into four regions from ventral to dorsal and calculated peak-to-peak times to estimate its propagation speed in both groups. Figure 5 E, G, H). The propagation speed of the wandering cores is similar in the unignited and ignited groups. Figure 5 F), the corresponding average velocity of the unlit rat was 0.117 mm / s, and the corresponding average velocity of the lit rat was 0.107 mm / s.

[0149] Although migratory hippocampal activity was similar in unlit and lit rat seizures, migration to the contralateral hippocampus was observed only in lit rats. In seizures with a migratory core, activity spreading from the iDHip to the cDHip was observed in 0% (0 / 8) of unlit seizures and 53% (8 / 15) (from 4 / 5 rats).

[0150] Although most rapidly generalizing epileptic seizures result in a wide distribution of activity ( Figure 3 F), but we found that in some seizures (three seizures from two lit rats), wandering hippocampal activity was the only significant response in the precortical pre-activated brain, suggesting that the core may play a key role in the generalization of seizures. Figure 5 I, J). The activity migrates from iVHip to iDHip (time points 2-5), then spreads to the ipsilateral cortex (6), the contralateral hippocampus (7-8), and finally to the contralateral cortex (9).

[0151] We have demonstrated whole-brain imaging of the dynamics of focal and FBTC seizures, as well as the underlying functional impairment networks present in these seizures. First, we developed a novel hippocampal ignition optogenetic model to reliably induce FBTC seizures. Then, we conducted a whole-brain study of the dynamics of the ventral hippocampal circuit using two stimulation modalities: short, mild 10 Hz stimulation to evaluate potential functional circuit changes, and long, intense 40 Hz stimulation to evaluate seizure circuit dynamics. We found that ignition led to chronic, widespread reorganization of the stimulated ventral hippocampal circuit, with the most significant changes in activity in the mePFC. Anxious behavior was exacerbated after ignition. Next, we imaged the whole-brain dynamics of individual focal and FBTC seizures to reveal their distinct propagation patterns, and we identified a slow-migrating core with high-amplitude activity that frequently accompanies seizures. Importantly, this core was the only active area in the brain before seizure generalization in three ignited FBTC seizures, suggesting that this core is a fundamental propagation mechanism for seizures and plays a crucial role in seizure generalization.

[0152] Our observation of persistent, widespread ignition-induced changes in the ventral hippocampal circuit demonstrates how specific circuits can functionally reorganize after repetitive aberrant activity. Therefore, it serves as a useful animal model for investigating the effects of recurrent epileptiform or seizure-like activity on underlying circuits. Furthermore, the reorganized circuit was associated with increased anxiety, suggesting the emergence of anxiety-sensitive circuits, one of the most common comorbidities of epilepsy. We found that the most significant changes in brain activity occurred in the mePFC, indicating an enhanced excitatory relationship between vHip and mePFC. Notably, these regions are two of the three nodes in the anxiety circuit, showing increased theta coupling upon animal entry into an anxiolytic environment, with associated anxiety behaviors subsiding when theta coupling was disrupted. Dysfunction of the vHip-mePFC circuit has been observed in patients with epilepsy and other hippocampal ignition models, and this circuit dysfunction is associated with cognitive deficits, another common comorbidity of epilepsy. In summary, these data suggest that developing therapies targeting vHip-mePFC circuit dysfunction could reduce the incidence of some of the most common comorbidities associated with epilepsy and enable effective treatment of conditions that may have a greater impact on quality of life than the seizures themselves.

[0153] The wandering hippocampal core, as a fundamental mechanism for seizure propagation, is of great significance for epilepsy surgery, requiring reliable localization of the seizure onset zone (SOZ). Our single seizure data correspond to a known SOZ, namely the ventral hippocampus, and these data demonstrate how the wandering core influences SOZ localization. Figure 3(D, F, and 5J). Using these data, we found that the SOZ may not be active throughout the seizure, the regional distribution of seizure activity can change rapidly, and the most active area is often not the SOZ. This has a direct impact on standard clinical SOZ identification methods, such as single-photon emission computed tomography (SPECT) and electrophysiology. SPECT uses a short-lived radioactive blood flow tracer that is manually injected during a seizure, so any injection delay can result in any activity in the SOZ being undetectable. Electrophysiology uses electrodes to map electrical activity during a seizure, and while it has excellent temporal resolution, location and orientation biases can occur that can lead to misclassification of the SOZ. An example of location bias is shown in single-seizure data where seizure activity propagates away from the electrode where the SOZ was implanted; although the seizure activity occurs elsewhere in the brain, the seizure is no longer detected on that electrode. Figure 3 (D, F, and 5J). Similarly, if electrodes are placed in the path of the wandering seizure core, the onset of seizure activity will be reflected as propagation rather than seizure. Therefore, developing methods capable of detecting the wandering core may help improve SOZ localization in epilepsy surgery.

[0154] The mechanism of the wandering hippocampal core remains unclear, and we do not know whether this phenomenon occurs outside the hippocampus. However, the intrinsic hippocampal circuitry is sufficient to support this core, as it has been observed in both unlit and lit rats. Seizure activity propagating at similar speeds has been reported in humans and animals when moving at approximately 0.1 mm / s, with ion diffusion and inhibitory restraint identified as potential key mechanisms. Interestingly, recent evidence suggests that slow-propagating activity is a major mechanism of sudden epilepsy-related death (SUDEP), requiring further investigation to determine whether this activity shares the same underlying mechanism as the wandering hippocampal core.

[0155] When an epileptic seizure propagates from the hippocampus, the activity follows known axonal pathways, thus involving synaptic mechanisms. Figure 5 J is an example of propagation via synaptic conduction, where seizure activity spreads from the ipsilateral hippocampus to the contralateral hippocampus. Synaptic seizure propagation mechanisms have been observed in mouse models of focal cortical seizures, where activity propagates to specific regions rather than non-selectively to adjacent areas. Developing therapies targeting these mechanisms could improve efficacy.

[0156] abbreviation. 'i' ipsilateral, 'c' contralateral, 'ChR2' channel rhodopsin 2, 'Amyg' amygdala, 'AudC' auditory cortex, 'CingC' cingulate cortex, 'DHip' dorsal hippocampus, 'DLThal' dorsolateral thalamus, 'EntC' entorhinal cortex, 'FrAssC' frontal symphysis cortex, 'Hypo' hypothalamus, 'InsulC' insular cortex, 'MDThal' dorsomedial thalamus, 'mePFC' medial prefrontal cortex, 'MotorC' motor cortex, 'OrFrC' orbitofrontal cortex, 'ParieC' parietal cortex, 'PiriC' piriform cortex, 'ReplC' posterior depressor cortex, 'Sept' septum, 'SomC' somatosensory cortex, 'Stria' striatum, 'TeAssC' temporal symphysis cortex, 'VHip' ventral hippocampus, 'VisC' visual cortex.

[0157] Materials and methods

[0158] Experimental Design. AAV-5-CAMKIIa-hChR2(H134R)-eYFP was injected into the right ventral hippocampus of 25 adult male Sprague-Dawley rats (Charles River Laboratories), and MRI-compatible electrodes were implanted to achieve simultaneous stimulation and electrophysiological recording. Simultaneously, MRI-compatible electrodes (REF Ben electrode paper and hippocampal paper) were implanted into the ipsilateral medial prefrontal cortex for electrophysiological recording. At least six weeks post-surgery, rats were imaged using simultaneous LFP optogenetic fMRI (ofMRI) to investigate changes in the hippocampal circuitry after ignition. Animals were randomly assigned to two groups: an experimental ignition group and a control group (n=12, n=13). At least one week after MRI scanning, the ignition group underwent ignition, and all rats were imaged again using LFP-ofMRI 1-2 weeks later. Approximately one month after LFP-ofMRI, a subset of rats (n=5 per group) were re-examined to evaluate the persistence of the ignition effect. Then, 10 weeks after LFP-ofMRI, evidence of depression and anxiety in the remaining animals (n=7, n=8) was assessed using a sucrose preference test, an open field test, and a forced swimming test to minimize the impact of acute seizures and evaluate the long-term consequences of the ignition. A behavioral test was performed one week later. At least one week after the completion of the behavioral test, seizures were imaged using LFP-ofMRI to investigate the nature of the seizure circuitry.

[0159] Viral injections and surgical procedures involving phototherapy and electrode implantation were performed. Animal husbandry and experimental protocols were strictly followed in accordance with guidelines from the National Institutes of Health (NIH) and the Institutional Animal Care and Use Committee (IACUC) of Stanford University. Animals were housed under controlled environmental conditions with a 12-hour light-dark cycle and free access to food and water.

[0160] In summary, rats were anesthetized with pure oxygen containing 5% isoflurane, and then isoflurane was maintained at a 2-3% supply throughout the procedure. 2 μl of AAV-5-CAMKIIa-hChR2(H134R)-eYFP was injected into the right ventral hippocampus (AP: -5.6 mm, LR: 5.7 mm, DV: 6 mm, with the dura mater as the base) using a 33-gauge needle connected to a Hamilton syringe. A constant infusion rate (150 nl / min) was maintained using an infusion pump (Micro 4, World Precision Instruments, FL). MRI-compatible carbon fiber electrodes (constructed using a 0.22 numerical aperture, 105 μm diameter step-index multimode fiber (ThorLabs, Newton, NJ)) were inserted as previously described (Duffy et al., 2015) with the electrode tip and fiber positioned directly above the injection site. Prior to implantation, the electrodes were checked to ensure a light transmission percentage greater than 80%, and the light transmission to the brain was assumed to be approximately that value. A single brass screw was inserted above the cerebellum to secure the dental cement and serve as both a grounding and reference electrode. A carbon fiber electrode was implanted into the right medial prefrontal cortex (AP: +3.24 mm, LR: +1.25 mm, DV: -3.4 mm, with the dura mater as the base point at a 10° angle). Finally, the electrode wires were soldered to a DF13 connector (Hirose, Japan), and all components were secured to the skull using light-cured dental cement. Buprenorphine sustained-release (1 mg / kg, sc) was administered preoperatively to reduce surgical pain and discomfort. Lidocaine (4%) and bupivacaine (0.25%) were also administered topically preoperatively and postoperatively. To allow time for virus-induced protein expression, experiments were conducted at least 6 weeks postoperatively.

[0161] Optogenetic ignition. Twelve rats were ignited using the following procedure, with continuous monitoring via synchronized video LFP recording in their cages. Animals were connected via fiber optic and electrical rotary connectors (Doric Lenses) to a 473nm (blue light) diode-pumped solid-state laser (Laserglow Technologies, Toronto, Canada) and a 16-channel BrainAmp ExG MR amplifier (Brain Products, Germany). Video was recorded using a Logitech C920 HD Pro webcam.

[0162] The afterfire threshold for each animal was first assessed by progressively increasing the stimulation intensity to induce electrocardiographic seizures in the absence of behavioral seizures (assessed using the Racine scale). The stimulation sequence was initiated at 1 mW for 10 seconds at 40 Hz (pulse width 7.5 ms), with power increasing by 1 mW each time, an interval between stimulations (ISI) of 1 minute, and a maximum power of 20 mW. If no afterfire was observed, the duration was increased by 2.5 seconds, and the power was reset to 1 mW.

[0163] Once the threshold is determined, ignition begins. The animal is stimulated a maximum of 12 times per day, or until a Grade 5 motor seizure occurs. Stimulation can be repeated every other day for a maximum of 12 days. Stimulation intervals are 15 minutes. If a Grade 5 motor seizure is observed within the first three stimuli of the day, the animal is considered ignited.

[0164] Using the same ignition criteria, ignition persistence was assessed in a subset of animals (five age-matched parallel control animals and five ignited rats) at 3 and 12 weeks after ignition by three stimuli (ISI of 15 minutes).

[0165] Simultaneous LFP-ofMRI data acquisition was used to assess ventral hippocampal connectivity. To assess ventral hippocampal connectivity, CAMKIIα cells in the ventral hippocampus were stimulated at 10 Hz using optogenetic stimulation and evaluated using simultaneous LFP-fMRI. Data acquisition was performed using a 7T horizontal drilling system (Bruker BioSpec 70 / 30) at the Stanford Center for In Vivo Imaging Innovation (SCi3). RF excitation was performed using a 2-channel volumetric coil with a diameter of 86 mm, and a 20 mm single-loop surface coil was used as the RF receiver. Rats were sedated by a bolus injection (0.1 mg / kg, sc) of dexmedetomidine, followed by continuous infusion (0.05 mg / kg, iv) via a cannula inserted into the lateral caudal vein. A single bolus injection of feraheme (15 mg / kg, iv) was used for cerebral blood volume (CBV)-weighted imaging to improve the contrast-to-noise ratio (Mandeville et al., 1998) and microvascular sensitivity (Zhao et al., 2006), an advantage over BOLD fMRI. Approximately 15 minutes after contrast agent injection, fMRI was acquired using a 4-segment spiral readout with the following acquisition parameters: TR = 0.75 ms, TE = 9 ms, flip angle = 30°, field of view = 32 x 32 mm, matrix = 70 x 70, slice thickness = 0.6 mm, number of slices = 30, number of repetitions = 130, and number of virtual scans = 4. For optogenetic stimulation, a block design was used, consisting of 5 s on and 55 s off, with stimulation at a pulse width of 10 Hz and 7.5 ms. At the end of each treatment, Atipamazole (0.5 mg / kg, sc) was administered to reverse the effects of dexmedetomidine. In addition to fMRI, anatomical (fast spin echo) scans were acquired to assess hippocampal volume with the following parameters: TR = 4755 ms, TEeff = 29.9 ms, RARE factor = 8, field of view = 30 x 30 mm, matrix = 256 x 256, and slice thickness = 0.6 mm. To assess hippocampal volume, hippocampal volumes were manually plotted on the ipsilateral and contralateral hemispheres of each animal. Baseline and post-ignition measurements were taken in both unignited and ignited animals, and animal volume was assessed by normalizing each hippocampal volume based on its baseline. For each hemisphere, a separate t-test was used to assess differences between the two groups to determine whether a significant reduction in hippocampal volume was associated with ignition.

[0166] Simultaneous LFP-ofMRI data acquisition was used to image the dynamics of the seizure circuit. To assess the dynamics of the whole-brain seizure circuit in unlit (n=7) and lit (n=5) animals, we used simultaneous LFP-fMRI and optogenetic stimulation of ventral hippocampal CAMKIIα cells. Imaging was not possible for three animals (unlit group n=1, lit group n=2) due to implant loss. Seizures were induced using a 40Hz pulse sequence (pulse width 7.5ms) at a predetermined post-fire threshold, as described above. Animals were paralyzed and ventilated for seizure LFP-ofMRI because motor seizures under dexmedetomidine sedation protocol resulted in significant movement. Anesthesia was induced with 5% isoflurane and maintained with 2-3% isoflurane, followed by cannulation of rats for drug infusion. Subsequently, animals were cannulated and injected with dexmedetomidine (0.1 mg / kg, sc). For the remainder of the procedure, isoflurane was slowly withdrawn, and dexmedetomidine (0.05 mg / kg, IV) and vecuronium (7.5 mg / kg) were continuously infused to maintain sedation and paralysis. Data acquisition was performed using the MRI system described above. For fMRI, data was acquired using a single EPI readout with the following parameters: TR = 1 s, TE = 16 ms, flip angle = 30°, field of view = 32 x 32 mm, matrix = 70 x 70, slice thickness = 0.6 mm, number of repetitions = 480, number of virtual scans = 4. Seizure induction began 90 s after the start of imaging. Atipamazole (0.5 mg / kg, SC) was administered at the end of each treatment to reverse the effects of dexmedetomidine.

[0167] Simultaneous LFP recording and processing. LFP recording was performed simultaneously with fMRI data acquisition using a 16-channel BrainAmp ExG MR amplifier (Brain Products, Germany), equipped with a 1000 Hz low-pass filter and a sampling frequency of 5000 Hz. Gradient artifacts were removed using principal component analysis, employing the method implemented by Liu et al. (Liu et al., 2012).

[0168] fMRI Preprocessing and Analysis. All fMRI data were preprocessed using SPM12, and 10Hz ventral hippocampal activity data were analyzed using SPM12. For preprocessing, the data were first smoothed at full width at half maximum (FWHM) using a 0.5mm Gaussian kernel, and motion was corrected using 6-parameter rigid registration. Images were manually masked, and then aligned with the common space using 12-parameter affine registration implemented in SPM. Signal convolution of stimulus blocks was performed using dual gamma-based group functions. Statistical activation maps were generated using a general linear model (GLM). For region analysis, the brain was automatically segmented using a general brain atlas, generating 44 regions, and then the activation volume and mean duration for each individual animal were calculated.

[0169] fMRI seizure propagation analysis. For each seizure on fMRI, seizure propagation was determined at the region of interest level. Scans were recorded into a brain atlas, low-pass filtered at 0.1 Hz, and segmented into individual regions using the aforementioned brain atlas. Mean durations were calculated using all voxels within a given region. Seizure onset time for each region was calculated by determining the time it took for the signal to reach a difference of four standard deviations from the baseline (60 s before stimulus onset), with the additional constraint that the signal must remain above the baseline for at least 5 s within the subsequent 10 s.

[0170] The relationship between hippocampal connectivity and seizure induction. To investigate the relationship between hippocampal connectivity and seizures induced in this network, we calculated the conditional probability that a region would be active during seizure induction, provided that it was active during a 10Hz non-seizure stimulus.

[0171] For 10Hz hippocampal connectivity data, the above procedure was used to generate a single subject voxel-level map, and thresholding was performed at P<0.001.

[0172] To determine the network associated with optogenetic seizure induction in each animal, we calculated the percentage of activated voxels during the first 10 seconds of induction. First, the percentage of maximum BOLD change was determined at the voxel level. Next, the images for each induction were binarized using the Otsu method, first applying a threshold that minimizes intraclass variance, and then the percentage of activated voxels during seizures within the animal was calculated to generate a graph that could evaluate network changes associated with optogenetic seizure induction. For group analyses, these graphs were averaged within groups to provide a network graph associated with optogenetic seizure induction.

[0173] Using a single-subject 10Hz hippocampal connectivity map and a single-subject seizure induction network map, we further evaluated the conditional probability of a region being active during seizure induction, provided it was active during 10Hz stimulation. First, all data were automatically segmented into corresponding data for each of the 44 regions using a 44-region brain atlas. Next, for both datasets, a 5% activation volume threshold was used to determine whether a brain region was active. Afterward, we calculated the conditional probabilities at the region level for the unlit and lit groups.

[0174] Sucrose preference experiment. Two bottles filled with fresh water were placed in the cages of individually housed rats (8 control rats and 7 flammable rats) for four days. The bottle positions were changed daily to allow the rats to adapt to the bottles and reduce their preference for a particular location. The bottles were weighed daily (9:00 AM to 10:00 AM). After the adaptation period, two bottles filled with fresh water—one full of water and the other full of 5% w / v sucrose solution—were placed in the cages. The bottles were weighed and their positions were changed daily (9:00 AM to 10:00 AM) for four days. The identities of the bottles were blinded during the study. Loss of interest was assessed by the daily water consumption and the ratio of water to sucrose solution.

[0175] Open field test. For this level of activity, the animals (8 control rats and 7 lit rats) were blinded and only unblinded after data processing. Rats were placed individually in custom-made chambers (1m x 1m x 0.3m) and recorded using a Logitech C920 HD Pro webcam. Automatic offline tracking was performed for the first five minutes of recording using Viewer3 software (Biobserve GmbH, Germany). Anxiety was assessed using distance traveled and time spent away from the edge of the activity area.

[0176] Forced swimming test. To this extent, the animals (8 control rats and 7 lit rats) were blinded and only unblinded after data processing. Rats were individually placed in custom-made cylindrical containers (30 cm in diameter, 90 cm in height) filled with water at 23–25°C to a depth of approximately 60 cm. A 10-minute pre-experiment was conducted the day before the experiment. Animal behavior was recorded for 5 minutes during the experiment using a Logitech C920 HD Pro webcam. Animals were dried with towels and returned to their cages. All data were analyzed offline. Immobility, swimming, and climbing were scored using a modified FST scoring system to assess dominant behaviors at 5-second intervals (Slattery and Cryan, 2012). Behavioral despair was assessed using an immobility score.

[0177] Statistical analysis. Values ​​are expressed as mean ± sem. If indicated, adjust p-values ​​using Holm's Bonferroni method for multiple comparisons. Perform statistical analysis using SPSS 21 or a custom MATLAB script. P < 0.05 is considered statistically significant.

[0178] Example 2

[0179] Understanding how dysfunctional brain networks contribute to and influence the spread and termination of seizures is a core objective of epilepsy research and can inform the development of targeted therapies for seizures and epilepsy-related comorbidities. In this study, we investigated the excitatory ventral hippocampal (VH) network before and one week after recurrent hippocampal seizures (ignition) (a procedure leading to chronic network changes), while simultaneously performing simultaneous electrophysiological and functional MRI with optogenetic stimulation on rats. We also directly imaged the whole-brain network dynamics of a single seizure to reveal focal and initial focal progression to bilateral tonic-clonic seizures. We also tested for two common epilepsy-related comorbidities: anxiety and depression. Finally, we report the relationship between the hippocampal network and the resulting seizure dynamics based on in-subject data. A widespread increase in excitatory VH network activity was detected after ignition. The largest changes were detected in the mPFC (LFP bandwidth increased 6.4-fold and activation volume increased 56% after ignition). Since the VH prefrontal cortex circuit is associated with anxiety and depression, we tested and found that lit rats were more anxious (n=7, 8, p=0.046). For seizure imaging, the duration of seizures in lit rats was 30.7±8.4 s longer than in unlit rats (p=0.001), and 15.5±2.2 more ROIs were activated (p<0.001). Propagation network analysis showed that activity in control animals remained ipsilateral, while activity in lit rats gradually spread to both cortices. mPFC activation was faster. The dorsomedial thalamus, a region associated with seizure generalization, was active only in lit rats. Next, we investigated the linear relationship between the VH network and the seizure network, finding that 95% of the variability in control animals was explained, while 77% of the variability in lit rats was explained (p<0.001, p=0.05, n=7, 5). Finally, we investigated the positive predictive value of VH network activity for seizure involvement, finding that in terms of consistently activated ROIs, the value was 0.86 ± 0.08 (3 ROIs, n = 7) in control animals and 0.89 ± 0.07 (14 ROIs, n = 5) in lithotripsy rats. This suggests that assessing VH activity can reliably reveal seizure pathways. In summary, these results reveal the whole-brain excitatory VH network that propagates hippocampal seizures and the long-term effects of recurrent seizures on these networks.

[0180] In this study, we investigated the excitatory ventral hippocampal (VH) network before and one week after recurrent hippocampal seizures (ignition) (a procedure leading to chronic network changes), while simultaneously subjecting rats to synchronized electrophysiological and functional MRI with optogenetic stimulation. We also directly imaged the whole-brain network dynamics of a single seizure to reveal focal and initial focal progression to bilateral tonic-clonic seizures. We also tested for anxiety and depression, two common epilepsy-related comorbidities. Finally, we report the relationship between the hippocampal network and the resulting seizure dynamics based on in-subject data.

[0181] Optogenetic ignition was performed. Ten out of twelve rats (83%) achieved ignition on day 7 of stimulation, and all rats ignited on day 11. The mean number of stimulations required for ignition was 53.6 (range: 29–117), the mean number of convulsive seizures (Racine grade 3–5) was 6.75 (range: 2–13), and the mean number of generalized seizures (Racine grade 5) was 2.9 (range: 1–9). Successful targeting of the ventral hippocampus with ChR2-eYFP was confirmed by fluorescence microscopy.

[0182] Ignition minimizes cell loss, so we tested whether any significant hippocampal volume loss could be detected by measuring hippocampal volume in anatomical MR scans obtained during the scanning process, before and after ignition, and in parallel control unignited rats. No significant difference in hippocampal volume was detected in either the ipsilateral or contralateral hippocampus at ignition compared to unignited animals. Figure 1 F, 0.63±0.82% vs 0.14±0.96%, p=0.7 and 1.37±1.22% vs -1.03±1.27%, p=0.19, lit vs unlit, in ipsilateral and contralateral hippocampi, respectively, mean ± sem), which supports the hypothesis that lit does not lead to significant cell loss.

[0183] To confirm the chronic effects of ignition, a subset of rats (ignited and unignited, n=5) were retested at 3 and 12 weeks post-ignition. All five ignited rats exhibited Racine grade 5 seizures at both time points, indicating that ignition persisted, while the unignited rats did not exhibit any obvious seizure behavior at either time point.

[0184] Local and whole brain effects on 10H zThe response of the ventral hippocampus to optogenetic stimulation (with and without ignition) and the effects of ignition on anxiety and depression. In unignited rats, 10 Hz optogenetic stimulation resulted in CBV activity primarily located in the ipsilateral hemisphere, dorsal and ventral hippocampus, septum, amygdala, and medial prefrontal cortex. In contrast, in ignited rats, the same stimulation resulted in activity extending beyond these areas and including regions in the contralateral hemisphere. When comparing the active volumes of different brain regions, we found that after multiple comparison correction using the Bonferroni-Holm method, the two groups of rats showed consistent volume differences in seven regions: ipsilateral frontal cortex (0.24±0.24% vs. 27.1±5.6%, p=0.004), temporal cortex (0.34±0.34% vs. 26.7±4.8%, p=0.0006), orbitofrontal cortex (1.4±0.7% vs. 23.8±3.9%, p=0.0005), insular cortex (0.2±0.18% vs. 13.0±2.7%, p=0.0047), and finally striatum (2.6±0.9% vs. 14.8±2.7%, p=0.011). There was no significant difference in the percentage of active ROI in the ipsilateral ventral or dorsal hippocampus between the two groups (17.3±3.0% vs. 27.8±3.1%, p=0.684, 10.1±2.9% vs. 24.4±5.9%, p=0.994).

[0185] Next, we investigated how active regions in unlit rats changed in lit rats by examining the amplitude of fMRI response to optogenetic stimulation. We did not detect differences in response amplitude in the ventral or dorsal hippocampus (unlit vs. lit, mean ± sem, p-values ​​adjusted using the Bonferroni-Holm method for multiple comparisons); VHip (2.1 ± 0.3% vs. 3.1 ± 0.4%, p = 0.122), dHip (0.8 ± 0.2% vs. 1.2 ± 0.2%, p = 0.178). However, significant differences were observed in the mePFC (1.3 ± 0.3%, 4.2 ± 0.6%, p = 0.0023), septum (1.6 ± 0.3%, 2.5 ± 0.3%, p = 0.082), and amygdala (1.1 ± 0.3%, 3.0 ± 0.4%, p = 0.0023).

[0186] Because we simultaneously measured local field potentials (LFP) in the iMePFC and iVHip using fMRI, we investigated whether complementary LFP responses existed in these regions. Upon ignition, the iMePFC showed a significant visual increase. We quantified this change by calculating the bandwidth response during the 5-s stimulation period (normalized based on the bandwidth 5 s before stimulation onset) and comparing the pre- and post-ignition responses with the corresponding responses in unignited rats. We did not detect any significant differences in the ventral hippocampus (F = 0.59, p = 0.81, n = 13 for unignited rats and 12 for ignited rats, analyzed using two-way repeated measures ANOVA). However, in the medial prefrontal cortex, we found a significant cohort-time interaction (F = 5.39, p = 0.031, n = 11 per cohort). Post-hoc analysis showed an increase of 6.4 ± 2.6-fold (mean ± sem) after ignition (paired t-test, t = 2.47, p = 0.033). This, combined with the fMRI response, indicates that ignition increases connectivity between the ventral hippocampus and the medial prefrontal cortex.

[0187] Since the ventral hippocampus-to-medial prefrontal cortex circuit is associated with anxiety expression, we investigated whether the ignition procedure leads to long-term behavioral disorders in some animals. We performed anxiety and depression tests, two of the most common epilepsy-related affective comorbidities. For depression, we conducted forced swimming and sucrose preference tests on the animals. Neither test showed a significant difference between unignited and ignited animals. For the forced swimming test, the immobility scores of unignited controls and ignited rats were 36.3 ± 0.99 and 34.1 ± 1.89 (mean ± sem), respectively, with a p-value of 0.303 calculated by a t-test. For the sucrose preference test, we found no evidence of a difference between unignited and ignited rats (F = 0.48, p = 0.499, two-way repeated measures ANOVA, mean estimates for unignited and ignited animals were 96.2 ± 0.9% and 83.9 ± 12.8%, respectively). When sucrose water was introduced during the experiment, the total fluid consumption increased in both groups (32.3 ± 8.0 ml in the unlit rat group, p = 0.005; 24.7 ± 6.9 ml in the lit rat group, p = 0.011, paired t-test). We found that lit animals spent less time in the center of the open field compared to unlit rats, suggesting that hippocampal ignition led to increased anxiety (12.5 ± 1.5% vs. 8.0 ± 1.4% of center time spent, p = 0.046, mean ± sem, unlit control rats vs. lit rats, t-test, n = 8 and 7). Importantly, the distances traveled by the two groups during the 5-minute test were similar (2446 ± 153 cm vs. 2203 ± 197 cm, p = 0.98).

[0188] Single seizures in unlit and lit rats were imaged using simultaneous LFP-fMRI. In ignited animals, induced seizures were associated with pronounced motion artifacts on MRI due to stereotyped behavior under dexmedetomidine sedation, reminiscent of behavior observed in awake ignited animals. Therefore, we developed a protocol based on dexmedetomidine sedation combined with vecuronium bromide (a short-acting neuromuscular blocking agent) to prevent movement in animals during seizure imaging.

[0189] In both lit and unlit animals, we used LFP to validate the occurrence of seizures following cessation of optogenetic stimulation in the ipsilateral ventral hippocampus (acquired simultaneously with fMRI). Importantly, the ipsilateral ventral hippocampal BOLD signal followed a similar trajectory to LFP, indicating that BOLD captured the seizure response. Examples of single seizures (achieved using LFP) in both lit and unlit animals, along with normalized BOLD changes corresponding to a single 1-second frame at a time point on LFP, are presented. In the seizures of the unlit animals, an initial burst of activity, reminiscent of a 10 Hz network, occurred in the ventral hippocampal circuitry, followed by activity propagating to the ipsilateral dorsal hippocampus, subsequently resulting in a negative BOLD response. Voxel-level maximum intensity projection (MIP) during the scan revealed a high-amplitude activity network.

[0190] In a single seizure in ignited animals, the area of ​​BOLD activity increased dramatically throughout the scan, and activity was now detectable in the cortex, indicating that the seizure generalized from the initial epileptic focus. Importantly, BOLD activity spread from the initial hippocampal network to the ipsilateral cortex and then to the contralateral cortex, demonstrating for the first time the dynamics of seizure propagation from focal to secondary generalized seizures across the entire brain. The MIP of the seizure summarized the high-amplitude activity of this experiment. Notably, for both unignited and ignited seizures, BOLD activity persisted until the end of the seizure, when it was detected on the LFP.

[0191] Seizure propagation in unlit and lit animals. Next, we quantified the propagation of BOLD activity in induced seizures in unlit and lit rats (n=20, from 7 rats; n=17, from 5 rats. One and two rats in the unlit and lit groups, respectively, lost their cephalopods and could not be used in this experiment). To capture propagation, we used automated segmentation atlases for region analysis, resulting in 44 ROIs. Examples of time series at the region level are shown below. Figure 4A. To determine propagation, we defined the regional seizure onset time as the time when activity first reached four standard deviations above the baseline 60 s prior to stimulus onset, and the activity had to remain above the threshold for 5 s within the subsequent 10 s. To validate that the procedure could distinguish between the unlit and lit groups, we compared the number of ROIs detected between the two groups and confirmed that the lit group had more consistently activated ROIs during seizures than the unlit group (23.4 ± 2.0 ROIs vs. 38.8 ± 1.0 ROIs, n = 20 vs. 17, control vs. lit, mean ± sem, p < 0.0001).

[0192] Next, we investigated the activity distribution of 44 regions in all unlit and lit seizures. Rearranging the radar plots along the central axis into mirror images of the two hemispheres reveals that seizures in lit rats were essentially always bilateral, while unlit rats also exhibited some activated regions, with these regions preferentially located in the ipsilateral hemisphere. To reliably estimate seizure onset time, we used only regions that were activated at least 80% in each seizure group (n=16–20 and n=14–17 in both unlit and lit seizures; seizure onset time calculated using 90% and 0% thresholds is shown in the supplementary plot).

[0193] The seizure onset times of the ROIs in each group were sorted from fastest to slowest, and color-coded for ipsilateral and contralateral hemispheres. In unlit animals, eight ROIs were estimated (iPiriC, IMePFC, iOrFrC, iVHip, iSept, iInsulC, iEntC, iTeAssC). Notably, all regions originated from the ipsilateral hemisphere. For lit seizures, activity was detected in 38 regions, notably with initial activation in the ipsilateral hemisphere followed by activation in the contralateral hemisphere. These ROIs included the eight found in unlit rats. Of the top 20 regions with the fastest seizure onset times, 18 were located in the ipsilateral hemisphere. Of the remaining 18 regions, 14 were located in the contralateral hemisphere. Next, we compared the seizure onset time of eight regions that were consistently activated in the unlit brain between the two groups. We found that the ipsilateral mPFC was activated faster in the lit brain than in the control group (4.1±0.34s vs. 2.8±0.26s, unlit control vs. lit, Holm-adjusted p = 0.044), while no differences were observed in other regions (Holm-adjusted p > 0.3): ipsilateral ventral hippocampus (4.2±0.43s vs. 4.4±0.76s). The durations of the transorbital cortex (4.2±0.51s vs. 3.0±0.21s), septum (4.6±0.62s vs. 7.1±1.34s), insular cortex (7.1±3.94s vs. 6.9±2.33s), piriform cortex (3.4±0.32s vs. 3.4±0.33s), entorhinal cortex (9.8±3.06s vs. 9.7±5.7s), and temporal cortex (10.8±2.14s vs. 8.6±3.46s) were 4.2±0.51s and 3.0±0.21s, respectively.

[0194] 10H z The relationship between the functional networks of ventral hippocampal stimulation and optogenetic seizure induction. We hypothesize that the regions of initial circuit activation during optogenetic seizure induction are similar to those activated by 10Hz ventral hippocampal stimulation, since both seizure- and non-seizure-inducing stimuli activate the ventral hippocampal circuits. To evaluate the relationship between the two networks, we compute the conditional probability of a region of interest (ROI) being active in the seizure-inducing network, provided it is active during 10Hz stimulation.

[0195] To estimate the seizure-inducing network for each seizure, we identified voxel-level maximum intensity values ​​during the initial 10 seconds of seizure induction. To determine voxel activity, images were binarized using the Otsu method, which minimized intra-class variance (mean threshold 2.49 ± 0.17% for the unlit group, 3.0 ± 0.19% for the lit group, mean ± sem, t-test, p = 0.32). Since each animal experienced 2–4 seizures, we calculated the activation frequency of each voxel and normalized it to the number of seizures for group comparison. Finally, we calculated the mean activation count of each voxel at the group level to generate the seizure-inducing network for comparison. Visually, there was a striking and remarkable similarity between the seizure-inducing network and the subthreshold stimulation network at the group level. We quantified the ROI activation levels in both groups using segmentation atlases and sorted the volumes of the lit group in descending order. After adjusting for multiple comparisons, we found no significant differences in any ROI, but this may be due to the small number of animals we had and the number of multiple comparisons involved. We investigated the number of active voxels between the two groups and found that the induction network in the lit rats had more active voxels than in the unlit rats, which supports the hypothesis that the seizure induction network was larger after ignition (3562±636 voxels in lit rats and 1784±347 voxels in unlit rats, t=2.64, p=0.025).

[0196] The varying vascular sensitivities to imaging procedures and stimulation modalities make it difficult to compare networks at the voxel level; therefore, we employed the ROI method to calculate conditional probabilities. For both the 10Hz stimulation network and the seizure-inducing network, we used a 5% activation volume threshold to determine whether an ROI was active. To examine the robustness of the model, we compared the number of activated ROIs in the 10Hz network and the optogenetic seizure-inducing network between the two experimental groups. Significant differences were detected between the groups (for the 10Hz network, unlit group = 5.6 ± 1.4 ROIs vs. lit group = 16.8 ± 2.6 ROIs, t = 4.10, p = 0.002; for the seizure-inducing network, unlit group = 20.6 ± 2.1 ROIs vs. lit group = 30.8 ± 3.4 ROIs, t = 2.69, p = 0.023).

[0197] We calculated the probability of an ROI being active in the seizure-inducing network, provided it was active in the corresponding 10Hz network in the individual animal, to investigate the relationship between the two networks. The conditional probability distributions were sorted from largest to smallest, and also from most consistent to least inconsistent based on their consistency of activation in the 10Hz network. In the unlit group, iVHip and iSept showed the highest consistency in activity in the 10Hz network, as did iMePFC and iAmyg in the seizure-inducing network. iMePFC and iAmyg were the second most consistent regions in activity in the 10Hz network (observed in 4 / 7 animals), and these ROIs were also active in subsequent seizure induction. In lit rats, iVHip, iMePFC, iOrbFrC, iFrAssC, iSept, iStria, iTeAssC, and iAmyg were active in the 10Hz network and also active in subsequent seizure induction.

[0198] Overall, the mean conditional probability for unlit animals was 0.76 ± 0.11, while the mean conditional probability for lit animals was 0.89 ± 0.05, indicating that regional activation in the 10Hz network was similar to that in the network activated during seizure induction 12 weeks after the early scan. Furthermore, this supports the hypothesis that early activity is associated with seizure induction, and subsequent activity is caused by activation of this network.

[0199] Following ipsilateral hippocampal activity propagation-induced seizures, we observed a striking phenomenon in the ipsilateral hippocampus of animal subgroups, where high-amplitude activity clusters migrated from the iVHip to the iDHip within seconds. We attempted to characterize this phenomenon in the seizures that demonstrated it (8 seizures in 4 animals from unlit rats and 15 seizures in 5 lit rats).

[0200] The iVHip LFP response overlapped with the BOLD signal from the iVHip, and the time points were highlighted by the corresponding fMRI. For fMRI, we highlighted only two slices from the iVHip and iDHip for visualization. In unlit animals, activity began at the iVHip and progressively moved upwards to the iDHip. The activity continued until the end of the electrocardiographic seizure. A similar pattern was observed in lit rats as the activity moved upwards to its peak. In this case, after reaching the iDHip, the activity appeared to cross the ipsilateral region to reach the contralateral DHip. These seizures were observed from different animals, with significant intrahippocampal propagation.

[0201] To further characterize intrahippocampal propagation, the ipsilateral hippocampus was divided into four regions, from the most ventral to the most dorsal. We quantified the peak-to-peak responses of the most dorsal and most ventral regions to estimate propagation time, finding a mean difference of 42.9 s (range: 0–66 s) in unlit rats and 46.9 s (range: 0–118 s) in lit rats. This translates to a propagation velocity of 0.07 mm / s–0.117 ms / s.

[0202] The materials and methods are as described in Example 1.

Claims

1. A method for analyzing events associated with a seizure in the brain, the method comprising: collecting electro-physiological data and magnetic resonance imaging (MRI) data simultaneously from an animal brain that has been kindled, the animal brain that has been kindled provides neural events that reflect changes in functionality and neural circuitry when subjected to a stimulus; and allows analysis of the neural events caused by a single seizure; analyzing the electro-physiological data and the MRI data to determine corresponding changes in the electro-physiological data and the MRI data indicative of a neural event, wherein analyzing the electro-physiological data comprises calculating the frequency band power during the stimulus divided by the frequency band power before the stimulus, and wherein analyzing the MRI data comprises: segmenting the MRI data of the animal brain into a plurality of individual regions; quantifying activation in the plurality of individual regions; and selecting a portion of the plurality of individual regions having a greatest activation; and generating a network graph associated with at least one neural event based on the MRI data, the electro-physiological data, or a combination thereof.

2. The method of claim 1, further comprising analyzing the MRI data to determine neural event propagation, the analyzing comprising: low-pass filtering the MRI data; registering the MRI data to a brain atlas; segmenting the MRI data into a plurality of individual regions based on the registration; and calculating an average time course of a plurality of voxels in an individual region of the plurality of individual regions.

3. The method of claim 2, wherein calculating the average time course comprises determining a time at which a signal reaches four standard deviations from a baseline.

4. The method of any one of claims 1-3, wherein the stimulus is an electrical stimulus and is directed to a region of interest.

5. The method of claim 4, wherein the region of interest is the hippocampus.

6. The method of any one of claims 1-3 and 5, wherein the neural events that reflect changes in functionality and neural circuitry comprise one or more of the following: focal progression to bilateral tonic-clonic FBTC seizures, excitatory ventral hippocampus (VH) network changes, subthreshold stimulus-triggered changes, and wandering seizure core induction.

7. The method of claim 6, wherein the stimulus that provides neural events that reflect changes in functionality and neural circuitry is an electrical stimulus.

8. The method of claim 7, wherein the electrical stimulus is delivered by optogenetic techniques.

9. The method of claim 8, wherein the animal brain comprises neurons genetically engineered to comprise a light-activatable polypeptide.

10. The method of claim 9, wherein the light-activatable polypeptide is a channelrhodopsin.

11. The method of claim 10, wherein the channelrhodopsin is operably linked to a promoter expressed in excitatory hippocampal neurons.

12. The method of claim 11, wherein the promoter is a calcium-calmodulin-dependent kinase II alpha (CaMKIIa) promoter.

13. The method of any one of claims 8-12, wherein the stimulus that provides neural events is below a threshold that triggers a seizure.

14. The method of claim 13, wherein the stimulation is 5 Hz to 15 Hz.

15. The method of claim 13, wherein the stimulation is applied to evaluate potential functional circuit changes.

16. The method of any one of claims 8-12, wherein the stimulation is sufficient to trigger a seizure.

17. The method of claim 16, wherein the stimulation is 35 Hz to 45 Hz.

18. The method of claim 16, wherein the stimulation is applied to evaluate seizure circuit dynamics.

19. The method of any one of claims 1-3, 5, 7-12, 14-15, and 17-18, further comprising identifying a wandering seizure core.

20. The method of claim 19, further comprising localizing a seizure onset zone (SOZ) from the identified wandering seizure core.

21. The method of claim 20, wherein the SOZ is imaged by single photon emission computed tomography (SPECT).

22. The method of any one of claims 1-3, 5, 7-12, 14-15, 17-18, and 20-21, wherein the kindling is achieved using electrical stimulation, optogenetic techniques, or chemical treatment.

23. The method of claim 22, wherein the optogenetic kindling is performed by: a. delivering a polynucleotide encoding a light-activatable protein to target neurons of the hippocampus, b. repeatedly illuminating the hippocampus with a frequency and pulse width sufficient to induce kindling; and c. repeating step b. over a period of days.

24. The method of any one of claims 1-3, 5, 7-12, 14-15, 17-18, 20-21, and 23, wherein the animal is sedated and given a short-acting neuromuscular blocker to prevent movement of the animal during imaging of the seizure.

25. The method of claim 24, wherein the animal is sedated with dexmedetomidine and blocked with vecuronium.

26. The method of claim 5, wherein the low-pass filtering is performed at 0.1 Hz.

27. The method of any one of claims 1-3, 5, 7-10, and 26, further comprising generating connectivity data by calculating a conditional probability that a single region of a plurality of regions of the brain of the animal was active during a neural event.

28. The method of claim 1, wherein generating a network map comprises: calculating a percentage of time during the induction period that a voxel was activated, wherein a maximum percentage BOLD change for that voxel is determined; and binarizing an image containing the voxel.

29. The method of claim 28, wherein binarizing comprises applying a threshold that maximally reduces intra-class variance.

30. The method of any one of claims 1-3, 5, 7-12, 14-15, 17-18, 20-21, 23, 25-26, and 28-29, wherein the network graph comprises a radar plot of a plurality of individual regions of an animal brain, wherein the plurality of individual regions are ordered from fastest to slowest in speed of neural event occurrence, and are color coded for ipsilateral and contralateral hemispheres of the animal brain.

Images