A Model of New-Onset Refractory Status Epilepticus (NORSE) Using Intrahippocampal Kainic Acid in Adolescent Mice

Abstract

New-onset refractory status epilepticus (NORSE) is a devastating neurological condition characterized by the sudden onset of prolonged, drug-resistant seizures in previously healthy individuals. A clinically relevant animal model is essential to elucidate its pathophysiology, improve treatment, and reduce morbidity and mortality in NORSE. We developed an adolescent mouse model of NORSE by inducing prolonged convulsive status epilepticus in postnatal day 40 (P40) C57BL/6J mice—corresponding to human adolescence—via unilateral intrahippocampal kainic acid (IHKA, 150 ng) injection. Progression of recurrent spontaneous electrographic seizures was monitored through depth electrodes on days 3, 5, 7, and 21 post-IHKA. Cellular and immune responses were examined using immunohistochemistry, Nissl staining, and flow cytometry. Behavioral outcomes were assessed using open-field and Y-maze tests. IHKA-injected mice developed spontaneous recurrent seizures by day 3 (18.7 ± 3.2 seizures/hour). Seizure frequency nearly doubled by day 7 (35.2 ± 4.1 seizures/hour) and continued through day 21 (39.7 ± 4.8 seizures/hour). Significant neuroinflammatory changes were observed, including >9-fold microglia activation, two-fold astrocytosis, cell death in the CA3 subfields, and monocyte infiltration into the hilus. Flow cytometry revealed a two-fold increase in CD8⁺ cytotoxic T cells and γδ T cells in the hippocampus. Behaviorally, IHKA mice exhibited increased anxiety and cognitive deficits by day 21. This adolescent IHKA model recapitulates hallmark features of NORSE—acute repetitive convulsive seizures with rapid progression to subclinical status epilepticus, marked immune activation, hippocampal injury, and behavioral impairment. Our murine model of NORSE provides a valuable platform for investigating NORSE pathogenesis and evaluating novel therapeutic strategies.

Introduction

New-onset refractory status epileptics (NORSE) is a life-threatening neurological disorder characterized by the sudden onset of refractory status epilepticus (RSE) in individuals with no prior history of epilepsy or other significant medical history. Although rare, with an incidence of 0.5–1.0 per 100,000 children per year, it carries devastating clinical consequences, including prolonged hospitalization, severe neurocognitive impairments, and high mortality rates (Gaspard et al., 2015). Conventional anti-seizure medications (ASMs) are largely ineffective, making the management of NORSE exceptionally challenging. In most cases, no structural, toxic, or metabolic etiology can be identified, and the pathophysiological mechanisms underlying NORSE remain poorly understood (Specchio and Pietrafusa, 2020).

Accumulating evidence suggests that neuroinflammation plays a central role in the onset and progression of NORSE and its subtype, febrile infection-related epilepsy syndrome (FIRES). Clinical and experimental studies have reported increased levels of proinflammatory cytokines and markers of immune activation in the brains of affected individuals (Choi and Koh, 2008; Howell et al., 2012; Iizuka et al., 2017). Moreover, infiltration of peripheral immune cells, particularly T-cells, into the central nervous system (CNS) has been observed in NORSE patients (Dilena et al., 2019), supporting immune dysregulation as a critical factor in disease pathogenesis (Dilena et al., 2019). Despite these findings, the absence of a reliable animal model that recapitulates the acute and chronic features of NORSE has posed a significant barrier to understanding its etiology and developing effective treatment.

Animal models have been indispensable for elucidating the underlying mechanisms of epileptogenesis and testing therapeutic interventions. Among these, the intrahippocampal kainic acid (IHKA) model in rodents is a well-established paradigm for inducing temporal lobe epilepsy (TLE), recapitulating hallmark features such as spontaneous recurrent seizures and hippocampal sclerosis (Bouilleret et al., 1999). Traditionally, IHKA has been employed in adult animals to model mesial TLE (mTLE), but accumulating evidence indicates that age is a key determinant of seizure phenotype, neuropathology, and disease outcome (Curia et al., 2008). While traditional adult IHKA models have been invaluable for understanding temporal lobe epilepsy (TLE), they do not adequately recapitulate the unique features of pediatric NORSE, including the age-dependent vulnerability, rapid seizure onset, and prominent immune activation. Adult IHKA models typically exhibit gradual epileptogenesis with lower seizure frequencies and develop hippocampal sclerosis over weeks to months (Bouilleret et al., 1999). In contrast, NORSE patients, particularly children and adolescents, experience sudden onset of severe, refractory seizures with rapid progression and prominent neuroinflammation. While neocortical injury models have also been developed to study time-dependent seizure spread and evolution of epilepsy post status epilepticus and neuronal loss (Kharatishvili et al., 2006; Singh et al., 2022), the hippocampal focus of our IHKA model is particularly relevant given the frequent involvement of temporal structures in NORSE patients. Therefore, an age-appropriate model is essential to capture the specific pathophysiology of pediatric NORSE.

In adolescent mice, the IHKA model yields distinct epileptic characteristics compared to adults, including a rapid onset of spontaneous recurrent seizures and accelerated hippocampal neurodegeneration (Levesque and Avoli, 2013). This susceptibility is likely attributable to developmental factors such as ongoing neurodevelopmental processes, enhanced synaptic plasticity, and maturing inhibitory networks (Koyama et al., 2012). Postnatal day 40 (P40) mice represent an adolescent developmental stage, corresponding to human adolescence based on metabolic maturation, synaptogenesis, and approaching sexual maturity (Semple et al., 2016). Given these age-dependent variations, we hypothesize that adolescent mice subjected to IHKA may exhibit seizure characteristics that model the severe epileptic phenotypes observed in pediatric and young adult NORSE and FIRES patients. Furthermore, the vulnerability of the adolescent brain to excitotoxicity and inflammation may render P40 mice particularly suitable for modeling the neuroinflammatory components of NORSE.

The goal of the present study was to establish a reliable rodent model of NORSE using adolescent (postnatal day 40; P40) mice, a developmental stage corresponding to human adolescence. By modifying the IHKA protocol, we aimed to induce severe, rapid-onset seizures that reproduce key clinical features of pediatric NORSE. We further sought to evaluate associated neuroinflammatory responses, including microglial and astrocytic activation and monocyte infiltration, which are implicated in seizure persistence and epileptogenesis (Varvel et al., 2016). Through evaluation of seizure phenotypes, hippocampal pathology, and neuroinflammatory markers, this study aims to establish a rodent model that recapitulates the clinical and neuropathological hallmarks of NORSE. Ultimately, this model may provide a valuable experimental platform for elucidating pathogenic mechanisms and testing novel therapeutic strategies targeting neuroinflammation in this catastrophic epileptic encephalopathy (Widmann et al., 2024).

Materials and Methods

Animals

All experiments and animal care procedures were conducted under the guidelines outlined by the Institutional Animal Care and Use Committee (IACUC) at Emory University and the University of Nebraska Medical Center. All experiments used C57BL/6J mice at postnatal day 40 (P40). Acute convulsive SE (Racine scale 3–5) was successfully induced in >95% of IHKA-injected mice across all cohorts. Mortality rates were comparable across cohorts (8–12% within 48 hours post-IHKA). Mice that did not develop observable SE or died within 48 hours were excluded from the study. All mice were purchased from Jackson Laboratory (Bar Harbor, ME), including (i) C57BL/6J mice and breeding pairs of (ii) B6.129P-Cx3cr1 ^tm1Litt/J (Cx3Cr1 ^GFP/GFP transgenic) mice and (iii) B6.129(Cg)-Ccr2 ^tm2.1Ifc/J (Ccr2 ^RFP/RFP transgenic) mice. To generate mice with dual fluorescent markers, Cx3Cr1^GFP/GFP mice were crossed with Ccr2 ^RFP/RFP mice, producing F1 offspring heterozygous for each fluorescent protein (Cx3Cr1 ^GFP/+ Ccr2 ^RFP/+). Despite heterozygosity, these mice maintained sufficient fluorescence intensity for a reliable detection of both cell populations. In this dual-reporter lineage, the Cx3Cr1 ^GFP/+ allele expresses green fluorescent protein (GFP) under the control of the Cx3Cr1 promoter, enabling visualization of brain microglia, while the Ccr2 ^RFP/+ allele expresses monomeric red fluorescent protein (RFP) as a knock-in replacement of the chemokine (C-C motif) receptor 2 coding sequence, allowing visualization of monocytes and their recruitment to active sites. All animals were housed in temperature- and humidity-controlled rooms with a 12 h light/dark cycle (lights on at 6:30 AM).

IHKA injection and implantation of electroencephalographic (EEG) electrodes

The intrahippocampal kainic acid (IHKA) injection was performed per our published method (Shen et al., 2015; Shen et al., 2022) with optimized KA dose (150 ng/ 150 nL saline) and stereotaxic coordinates for adolescent P40 mice (AP −1.7 mm, ML −1.7 mm, DV −1.8 mm, relative to bregma) targeting the Cornu Ammonis 1 (CA1) region. An equivalent volume (150nL) of saline was injected into the Sham control animals. During the same procedure, EEG electrode implantation was performed as in our previously published method (Shen et al., 2022) to record intrahippocampal and cortical electrographic seizures. Specifically, one cortical electrode (AP −0.9 mm and ML - 0.3 mm ) and a bipolar intrahippocampal electrode (AP −1.7 mm, ML −1.7 mm, DV −1.8 mm) were installed and secured with dental cement on the skull. After surgery, the mice were continuously observed for 3 hours. A range of irregular behaviors were observed in mice post-IHKA: Initial behaviors included brief pauses with a stiff, tonic posture characterized by rigid whiskers and a tonic tail, followed by intermittent repetitive spinning behavior that lasted over an hour. Additional observed behaviors included clonic jerks, tonic posture, loss of balance, sudden jumps, and generalized tonic-clonic seizures. Seizures persisted for approximately 2 to 3 hours post-surgery.

Electroencephalogram recording and analysis

Electroencephalograms (EEG) were performed according to previously published methods (Shen et al., 2015; Shen et al., 2022). Mice were singly housed while tethered for the acquisition of the EEG recordings, with free access to water and food. At 3, 5, 7, and 21 days after IHKA injection, each mouse underwent an EEG evaluation of its epileptic phenotype, which included a 24-h period of EEG monitoring and recording (P511/P122 Grass Instruments, Astro-Med, West Warwick, RI) (Fig. 1A). Electrical brain activity was digitized (ML880 PowerLab 16/30; AD Instruments, Colorado Springs, CO), and quantification of EEG seizures was determined as in our previous work. Electrographic seizures were defined as high-voltage rhythmic EEG activity lasting 5 seconds or longer (Shen et al., 2015; Shen et al., 2022). Seizure burden was calculated as the percentage of total recording time spent in seizure activity (i.e., [total seizure duration/total recording time] × 100%).

Figure 1. Progression of electrographic seizures following IHKA injection at P40.

(A) Experimental timeline of IHKA injection, EEG implantation, and subsequent behavioral assessments. (B) Representative EEG traces 7 days post-injection. Upper trace (blue): SHAM (vehicle-injected control) showing normal background activity. Lower trace (red): NORSE (kainic acid-injected) showing clustered seizures, with the magnified view depicting the boxed region. Quantification of recurrent seizure frequency (seizures/hour) shown as group mean (C) and individual animal trajectories (D) across days 3, 5, 7, and 21 post-IHKA. Percentage of time spent in seizures, represented as group mean (E) and individual animal data (F) across the same timepoints. IHKA-injected (NORSE) mice displayed significantly elevated seizure frequency and burden compared to sham controls (n = 8 per group). Data are presented as mean ± SEM. Statistical analysis: unpaired t-test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Immunohistochemistry and Histology

To assess biochemical and cellular changes in the adolescent IHKA-induced NORSE model, a separate cohort of mice was sacrificed seven days post-surgery (i.e., P47) following the completion of EEG recordings. Age-matched vehicle-injected sham controls were processed in parallel. After transcardial perfusion with 4% formaldehyde, brains were post-fixed, cryoprotected in 30% sucrose-PBS, and sectioned into 30 μm coronal slices using a freezing microtome (Microm, GmbH, Walldorf, Germany). For IHC, sections were pretreated with citrate buffer (80°C, 30 min), blocked in blocking buffer (2% goat serum, 0.05% Triton X-100, 1% BSA), and incubated with 1% sodium tetraborate. Primary antibodies, anti-glial fibrillary acidic protein (GFAP) (AM020, 1:200, Biogenex) and anti-ionized calcium-binding adaptor protein-1 (IBA1) (ab5076, 1:200, Abcam), were applied at 4°C for 48 h to target astrocytes and microglia, followed by HRP-conjugated secondary antibodies for 90 min at room temperature, and DAB-nickel visualization (Vector Labs, Burlingame, CA). For immunofluorescence, sections were blocked in blocking buffer (2% donkey serum, 0.05% Triton X-100, 1% BSA), incubated with anti-neuronal nuclei (NeuN) (MAB-377, 1:500, Millipore Sigma) at 4°C for 24 h, and visualized using fluorochrome-conjugated secondaries and DAPI mounting (Shen et al., 2020). Nissl staining was also performed to evaluate hippocampal cellular damage, with shrunken, hyperchromic cells indicating neuronal loss (Sarlaki et al., 2022). Quantification of optical density and degenerating cells confirmed neuropathology in IHKA-injected mice. Three independent brain sections from each brain were processed in parallel with controls omitting primary or secondary antibodies, using identical solutions and incubation times.

Image Quantification of Densitometry

High-resolution digital images were acquired using the LasX software system (Leica, Buffalo Grove, IL, United States). Optical intensity analysis was performed using ImageJ software (ImageJ, US. National Institutes of Health, Bethesda, MD; ImageJ.nih.gov/ij/). All image processing was applied identically across experimental groups. Immunoreactive material was measured in the same designated area of CA1 and/or CA3 for each sample and expressed as relative optical density (ROD) per area, with the control group serving as baseline. Three levels were measured for each mouse, and data analysis is expressed as the mean ± SEM of ROD. The average levels in the treatment groups were normalized to those in the corresponding control group (as the baseline).

Lymphocyte Flow Cytometry

A cohort of mice was sacrificed two weeks after IHKA (n = 11) or Sham (n = 10) injections and perfused with PBS; then, the individual hippocampi were dissected. For flow cytometry assessment, mouse hippocampal tissues were pooled into two groups, i.e., the ipsilateral hippocampus of IHKA- and sham-injected mice. Lymphocytes were then isolated from the tissue following our protocol described previously (Xu et al., 2018). Cells were incubated with mouse Fc-block (#553142, BD Biosciences) to prevent nonspecific binding, followed by appropriate antibody cocktails for cellular surface marker staining. Antibodies were sourced from BD Biosciences (CD11b/V500, #562128) and BioLegend (γδTCR/APC, #118115; CD3/FITC, #100203; CD4/BV421, #100443; NK1.1/PerC-Cy5.5, #108727; CD8/Alexa700, #100729). To identify dead cells, Live/Dead Fixable Blue (L-23105; Invitrogen) was added before fixation with 1% formaldehyde in PBS. Each incubation step was conducted at 4°C for 30 minutes in the dark. Cell populations were finally enumerated using a six-laser Fortessa flow cytometer (BD Biosciences), and the data were analyzed with FlowJo software (TreeStar). Gating strategy for identification of leukocyte populations was as follows: Initial gating on forward scatter (FSC) and side scatter (SSC) to identify leukocytes, followed by exclusion of dead cells using Live/Dead staining. CD3+ T cells were then identified, and further subdivided into CD4+ T cells, CD8+ T cells, and γδ T cells based on expression of CD4, CD8, and γδTCR, respectively.

Behavior evaluation

The open-field test was used to assess locomotor activity and anxiety-like behavior in IHKA-injected adolescent mice, based on our previously published method (Wang et al., 2013) with modifications. Mice were individually placed in a 25 × 25 cm arena with 35 cm-high walls for a 10-minute session, providing a novel environment in which to assess exploratory behavior and anxiety responses. Movement was recorded via a video camera and analyzed using ANY-maze software (Stoelting Co., Wood Dale, IL). Quantified parameters included: (i) locomotor activity, measured as total distance traveled and average speed; and (ii) exploratory behavior, assessed by the distance traveled in the center zone, defined as the area at least 10 cm from the arena walls. Increased time or distance spent in the center zone indicates reduced anxiety, whereas center avoidance reflects heightened anxiety-like behavior (Prut and Belzung, 2003). Spatial recognition memory was evaluated using the Y-maze novel arm test (Lalonde, 2002), which included a habituation (sample) phase and a test phase. During the sample phase, one arm of the Y-maze was blocked, and mice were allowed to explore the start and familiar arms for 3 minutes. Following a 30-minute inter-trial interval, the test phase was conducted with all three arms accessible. Behavioral parameters, including time spent in each arm, number of entries, and latency to enter the novel arm, were recorded and analyzed using ANY-maze software. A preference for the novel arm, reflected by increased exploration time or entries, is indicative of intact spatial recognition memory in healthy mice.

Statistical Analyses

All data were analyzed using GraphPad Prism software. Normality of data distribution was assessed using the Shapiro-Wilk test. The quantitative data are presented as mean ± SEM and were analyzed using unpaired t-tests for comparisons between two groups (Sham vs IHKA). For behavioral outcomes, two-way ANOVA (sex × treatment) was performed to assess sex-dependent effects. For longitudinal seizure frequency data across multiple time points, repeated measures analysis was performed. Specific statistical tests used for each comparison are indicated in the figure legends. A p-value < 0.05 was considered significant.

Results

Evolution of epilepsy in IHKA-injected adolescent mice

To establish a preclinical NORSE model, we utilized adolescent (P40) C57BL/6J mice and optimized the dosage of IHKA injection to balance seizure induction and survival. In pilot testing, a KA dose of 200 ng (1 mg/mL in saline) induced status epilepticus in 80% of injected mice, but resulted in 50% mortality. Reducing the dose to 150 ng maintained a high SE induction rate (73%) while reducing mortality to ~ 20%. All subsequent experiments in the present study were therefore conducted using 150 ng KA.

Following IHKA injection, all mice developed acute convulsive status epilepticus characterized by continuous behavioral seizures (Racine scale 3–5) lasting 20–30 minutes, followed by intermittent motor seizures over the subsequent 4–6 hours. By 24 hours post-injection, overt convulsive seizures subsided, but EEG recordings from day 3 onward revealed continuous electrographic seizures consistent with subclinical status epilepticus. Following IHKA injection (Fig. 1A), mice displayed an evolving sequence of seizure behaviors: first pausing and exhibiting intermittent tonic posturing of limbs, rigid whiskers, and tonic tails; then developing repetitive spinning behavior for over an hour, followed by clonic jerks, loss of balance, sudden jumps, or generalized tonic clonic seizures. After recovery from EEG implantation and IHKA-induced SE for two days, subclinical seizures were first detected on day 3 post-IHKA, whereas no epileptiform activity was observed in sham controls (n = 8/group; Fig. 1B). IHKA-injected NORSE mice developed multiple hourly seizures (Fig. 1B) on day 3 (18.7 ± 3.2 seizures/hour), progressively increasing on day 5 (32.0 ± 4.1 seizures/hour), and on day 7 (35.2 ± 4.1 seizures/hour), and persisting into day 21 (39.7 ± 4.8 seizures/hour) (Fig. 1C–D). Consistent with the seizure frequencies, the percentage of time spent in seizure (seizure burden) also rose over three weeks, indicating ongoing seizure activity and progressive epileptogenesis in this model. (Fig. 1E–F). Average electrographic seizure duration increased from Day 3 (13.2 ± 1.2 sec) to Days 5, 7, and 21 (19.5 ± 0.8, 18.6 ± 2.2, and 20.4 ± 1.5 sec, respectively; Supplementary Figure S1). Analysis of 1,455 individual seizures revealed that 66% were brief (5–20 seconds), 15% were moderate duration (20–30 seconds), and 19% were prolonged (>30 seconds). Behavioral monitoring during EEG recordings revealed that most electrographic seizures were subclinical, with only occasional behavioral manifestations (head bobbing, arrest) observed during longer seizure events.

Hippocampal immune activation

To evaluate hippocampal neuroinflammation, we performed immunohistochemistry targeting markers of microglia (IBA-1) and astrocyte (GFAP) and utilized Cx3Cr1 ^GFP/+ Ccr2 ^RFP/+ reporter mice to distinguish resident microglia from infiltrating monocytes. Compared to shams, IHKA mice showed markedly increased microglial activation, evidenced by significantly increased IBA1 immunoreactivity at day 7 in both CA1 (p < 0.0001) and CA3 (p = 0.0004) (Fig. 2A–F). Astrocyte activation was also significantly increased in IHKA-injected mice, with enhanced GFAP staining observed in CA1 (p = 0.0001) and CA3 (p = 0.0086) (Fig. 2G–L).

Figure 2. Microglial and astrocytic activation in the hippocampal CA1 and CA3 regions 7 days after IHKA injection.

Representative images of IBA-1 immunoreactivity in CA1 (A, B) and CA3 (C, D) regions of IHKA-induced (NORSE) mice compared with sham controls (SHAM). Corresponding GFAP immunoreactivity in CA1 (G-H) and CA3 (I-J). Quantification of IBA-1 (E, F) and GFAP (K, L) positive-stained areas in the CA1 and CA3, respectively. Data are presented as mean ± SEM. Statistical analysis: unpaired t-test. **p < 0.01, ***p < 0.001, ****p < 0.0001 versus sham. Scale bar: 100 μm.

Analysis of Cx3Cr1 ^GFP/+ Ccr2 ^RFP/+ reporter mice further revealed concurrent activation of resident microglia (GFP⁺) and infiltration of peripheral monocytes (RFP⁺) in the hilus at day 7 post-IHKA (Fig. 3B–C). The hilus was selected for quantification because it contains a dense vascular network and is susceptible to blood-brain barrier (BBB) disruption and immune cell entry (Varvel et al., 2016). Sham mice exhibited ramified, resting microglia and negligible RFP signal in both ipsilateral and contralateral hippocampi (Fig. 3A). In contrast, IHKA-injected reporter mice displayed sustained GFP⁺ microglial activation in bilateral hippocampi, with greater activation ipsilateral to the injection site) (Fig. 3B), consistent with the IBA-1 staining pattern shown in Fig. 2. Within the hilus (Fig. 3B, white boxed area), robust accumulation of RFP⁺ monocytes (white arrows) was observed alongside activated microglia (yellow arrows) (Fig. 3C) . Quantitative analysis confirmed a significant increase in RFP⁺ monocytes in the hilus of IHKA-injected hippocampi compared to sham (Fig. 3D), indicating BBB disruption and recruitment of peripheral immune cells. These findings suggest that BBB disruption and subsequent monocyte infiltration may contribute to the neuroinflammatory environment characteristic of NORSE.

Figure 3. Microglial activation and monocyte infiltration in the hippocampal hilus.

Immunofluorescence imaging of GFP+ microglia in the contralateral and ipsilateral hippocampus of sham (A) and IHKA mice (B) 7 days post-injection. Sham controls exhibited resting GFP+ microglia, whereas IHKA mice showed robust GFP+ microglia activation in both hippocampi, more pronounced on the ipsilateral side. White boxes indicate the hilus regions selected for monocyte infiltration analysis. (C) RFP+ monocytes (white arrows) and merged images showing colocalization with GFP+ microglia (yellow arrows) within the hilus of IHKA mice. (D) Quantification reveals a significant increase in RFP+ monocytes in IHKA injected (NORSE) mice versus controls (SHAM). Data are presented as mean ± SEM. Statistical analysis: unpaired t-test. (E) Flow cytometry analysis of lymphocyte counts in pooled hippocampal samples from IHKA (NORSE) versus SHAM mice. Ipsilateral hippocampi from 10–11 mice were pooled per group for analysis. Data represent absolute cell counts from pooled samples. Scale bars: 500 μm (A, B); 100 μm (C).

Due to the limited volume of hippocampal tissue from individual mice, samples were pooled (10–11 ipsilateral hippocampi per group) for flow cytometry analysis. While this approach precludes assessment of inter-animal variability, it provides robust quantification of overall immune cell infiltration patterns. Flow cytometry of pooled hippocampal tissue revealed these findings, showing twofold increases in CD8⁺ cytotoxic T cells and γδ (GD) T cells in the hippocampus of IHKA mice (Fig. 3E). Together, these data indicate pronounced innate and adaptive immune activation in the hippocampus of this adolescent IHKA model of NORSE.

Neurodegenerative Changes

Neuronal cell loss and damage were evaluated by Nissl staining and NeuN immunofluorescence. Nissl staining revealed shrunken, hyperchromic neurons - hallmarks of cellular degeneration (Sarlaki et al., 2022) (Fig. 4A). Quantitative analysis showed a significant reduction in Nissl-positive cells (p = 0.0004) and an increased percentage of damaged cells (p = 0.0030) in the CA3 region of IHKA-injected mice (n = 10) compared to sham controls (n = 7) (unpaired t-test) (Fig. 4B–C). NeuN immunostaining confirmed a marked decrease in CA3 NeuN-positive neurons in IHKA-injected mice versus sham controls (p = 0.0041) (Fig. 4D–E). These findings indicate prominent hippocampal neurodegeneration in this adolescent IHKA model.

Figure 4. Neurodegenerative changes in the hippocampus 7 days after IHKA.

(A) Representative Nissl-stained images showing neuronal damage and cell loss in the CA3 region. Damaged neurons appear darker with irregular borders (red arrows). Quantification of Nissl-positive cell numbers (B) and percentage of damaged cells (C) in the CA3, comparing IHKA-injected (NORSE) mice with sham controls (SHAM). (D) Representative NeuN and DAPI immunofluorescences in the CA3; yellow box indicates the analyzed regions for NeuN. (E) Quantification of NeuN-positive cells expressed as relative optical density (ROD) versus sham controls. Data are presented as mean ± SEM. Statistical analysis: unpaired t-test. **p < 0.01, ***p < 0.001. Scale bars: 200 μm (A, upper); 100 μm (A, lower); 150 μm (D).

Behavioral impairments in the NORSE model

To assess behavioral outcomes, a separate cohort of mice was tested in the open-field and the Y-maze paradigms, during the light phase, at 1 and 3 weeks post-IHKA. In the open-field test, IHKA-injected mice (n = 7) exhibited increased thigmotaxic (wall-hugging) behavior without changes in locomotion, as total distance and average speed were comparable to controls (n = 9), indicating no locomotor impairments (Fig. 5A–B). However, at one week post-IHKA, IHKA-injected mice spent significantly less time in the center zone compared to sham controls (p = 0.0225, unpaired t-test), consistent with increased anxiety-like behavior (Fig. 5C). At 3 weeks, Y-maze testing revealed significant spatial memory deficits: IHKA-injected mice spent less time (p = 0.0391) and made fewer entries (p = 0.0006) into the novel arm compared to the shams (Fig. 5E–F). No significant sex-dependent differences were observed in behavioral outcomes (Supplementary Table S1). Male and female mice performed similarly and responded similarly to IHKA treatment. Behavioral outcomes were analyzed by sex using two-way ANOVA (sex × treatment). No significant sex main effects were observed (open field week 1: F(1,11) = 0.048, p = 0.830; Y-maze time in novel arm week 3: F(1,10) = 0.068, p = 0.799; Y-maze novel arm entries week 3: F(1,11) = 0.513, p = 0.489). No significant sex × treatment interactions were detected (open field week 1: F(1,11) = 2.168, p = 0.169; Y-maze time week 3: F(1,10) = 0.364, p = 0.560; Y-maze entries week 3: F(1,11) = 0.196, p = 0.667). Together, these results indicate that IHKA-induced status epilepticus in adolescent mice produces both early anxiety and progressive cognitive impairments.

Figure 5. Behavioral impairments in IHKA-injected adolescent mice.

(A) Total distance traveled and (B) average movement speed in the open-field test, showing no significant locomotor deficits in IHKA-injected (NORSE) mice compared to sham controls. (C) Representative open field traces. (D) Percentage time spent in center zone, indicating increased anxiety-like behavior in IHKA-injected (NORSE) mice. Y-maze test showing the percentage of time spent (E) and percentage of entries into the novel arm (F), demonstrating impaired spatial memory in IHKA-injected mice. Data are presented as mean ± SEM. Statistical analysis: unpaired t-test. *p < 0.05, ***p < 0.001

Discussion

NORSE is defined as ‘a new onset of refractory status epilepticus (SE) without a clear acute or active structural, toxic, or metabolic cause in a patient without active epilepsy or other preexisting relevant neurological disorders (Hirsch et al., 2018; Wickstrom et al., 2022b). In establishing a rodent model of NORSE, we sought to align closely with this clinical definition. We used wild-type C57BL/6J adolescent mice with no preexisting neurological symptoms, spontaneous seizures, or structural/metabolic abnormalities before IHKA injection. Within 2–3 hours post-IHKA injection, mice exhibited intermittent escalating motor seizures, including tonic tail extension, spinning, jerking, running, and popping, culminating in generalized tonic-clonic seizures (Ndode-Ekane and Pitkanen, 2013; Twele et al., 2016). Following acute intermittent convulsive SE, mice developed spontaneous recurrent electrographic seizures that evolved into continuous nonconvulsive SE by day 3 post-IHKA—closely resembling the clinical course seen in NORSE patients who experience continuous, relentless, subtly clinical, electrographic seizures. Histological and cytological analyses revealed robust glial activation, monocyte infiltration, and elevated CD8 and γδ T-cell populations, suggesting pronounced both innate and adaptive immune activation. Behaviorally, IHKA-injected mice exhibited deficits, including increased anxiety-like behaviors and impaired cognitive functions. Collectively, these electrographic, immunopathological, and behavioral features indicate that IHKA in adolescent mice effectively models key aspects of NORSE and provides a relevant platform for preclinical evaluation of disease mechanisms and therapeutic strategies.

Distinct epileptic phenotypes in adolescent vs. adult IHKA mice: insights from the NORSE model

IHKA produces varied epileptic phenotypes in adolescent versus adult mice. The IHKA model has long been recognized as a robust model of human mesial temporal lobe epilepsy (mTLE), the most prevalent focal epilepsy in adults (Falconer et al., 1964; Margerison and Corsellis, 1966; Berkovic et al., 1991; Levesque and Avoli, 2013). mTLE is pathologically defined by progressive neuronal loss and gliosis, resulting in hippocampal sclerosis and selective neuronal degeneration, particularly within the CA1 and CA3 regions of the hippocampus (Berkovic et al., 1991; Riban et al., 2002). Clinically, mTLE typically follows a latent period between the initial insult (such as traumatic brain injury or convulsive status epilepticus, often febrile) and the onset of spontaneous seizures.

In adult mice (≥ 6–8 weeks old), IHKA induces hippocampal paroxysmal discharges and electrographic seizures within weeks of KA administration, stabilizing to a plateau after several weeks and months (Riban et al., 2002). The model reproduces key features of human mTLE, including a latent period preceding the development of spontaneous recurrent seizures, hippocampal sclerosis, and pharmacoresistance (Berkovic et al., 1991; Bouilleret et al., 1999; Riban et al., 2002). Notably, adult IHKA mice exhibit resistance to several commonly used antiseizure medications, reinforcing their value for evaluating new therapeutic approaches (Widmann et al., 2024).

In adolescent (P40) mice, IHKA injection induces acutely developed convulsive SE, followed by spontaneous recurrent nonconvulsive electrographic seizures detected by day 3. While continuous EEG monitoring from the time of SE was not performed in this study, the early detection of seizures at day 3 is consistent with reports of rapid seizure onset in younger mice (Levesque and Avoli, 2013; Tse et al., 2014). This repid progression likely reflects increased neuroplasticity and ongoing neurodevelopmental remodeling characteristic of the adolescent brain (Dube et al., 2010; Koyama et al., 2012). Adolescent animals also exhibit greater sensitivity to KA and a consistent convulsant response (Curia et al., 2008; Schauwecker, 2011), resulting in severe acute seizures. Consistent with our findings, IHKA-induced seizures in younger mice produce extensive hippocampal neurodegeneration and mossy fiber sprouting (Curia et al., 2008; Umpierre et al., 2014). These results highlight the critical influence of age on seizure dynamics and underscore the need to account for developmental stages when interpreting IHKA data (Curia et al., 2008; Koyama et al., 2012).

Clinically, NORSE is defined by its acute presentation (Hanin et al., 2024b) while TLE is characterized by chronic anatomical pathology. The relationship between NORSE and TLE represents distinct yet interconnected aspects of epilepsy. Emerging clinical evidence suggests that NORSE can evolve into TLE-like pathology, as patients recovering from refractory SE frequently develop hippocampal sclerosis and chronic mTLE (Gordon Boyd et al., 2012). Transcriptomic analyses have revealed shared molecular signatures, including upregulation of excitatory neuronal markers and genes linked to microglial activation and inflammasome signaling (Hanin et al., 2024a). In line with these findings, our adolescent IHKA model captures this intersection between neuroinflammation and epileptogenesis, bridging the gap between the NORSE and TLE spectrums. The seizure patterns observed in our adolescent IHKA model share characteristics with seizures in adult IHKA models that recapitulate human temporal lobe seizure onset patterns (Lisgaras and Scharfman, 2022). Specifically, we observed predominantly focal seizures with hippocampal onset, similar to the patterns described in IHKA model in adult mice and in human mesial temporal lobe epilepsy (mTLE). This is particularly relevant as many NORSE survivors develop chronic temporal lobe epilepsy. The progression from acute convulsive SE to chronic subclinical electrographic seizures in our model may reflect the transition from acute NORSE to chronic epilepsy observed in human patients, providing a platform to study both the acute and chronic phases of the disease.

These observations are consistent with published reports of age-dependent trajectories in the evolution of epilepsy in IHKA models. Administration of KA into the hippocampus of P40 mice led to early seizure onset (within 3 days), whereas a latent period of one week to one month before the onset of the first spontaneous seizures was observed in adult mice and rats (Levesque et al., 2015). While we did not perform direct age comparisons in the current study, previous literature indicates that adult mice display more gradual epileptogenesis with pronounced hippocampal sclerosis (Champsas et al., 2024). Future studies directly comparing adolescent and adult responses would be valuable to fully characterize age-dependent differences in NORSE phenotypes in IHKA models. Employing IHKA across developmental stages thus provides a powerful framework for linking clinical syndromes—such as NORSE and mTLE—to their shared mechanistic underpinnings, enabling the rational design of age-specific and disease-stage–targeted therapeutic strategies.

Our adolescent IHKA model shares some features with traditional adult IHKA models, including focal hippocampal injection, subsequent spontaneous recurrent seizures, neuroinflammation, and hippocampal sclerosis. However, several key differences distinguish our model and position it as specifically relevant for NORSE research: (1) Age specificity: P40 mice correspond to human adolescence and young adulthood, matching the epidemiological peak of NORSE; (2) Seizure phenotype: Our model exhibits very high-frequency electrographic seizures (20–40/hour) from day 3, in contrast to the lower frequency (2–5/hour) reported in adult models; (3) Immune response: We observed prominent T-cell infiltration (CD8+ and γδ T cells), which is particularly relevant to NORSE immunopathology; (4) Acute-to-chronic transition: The rapid progression from acute convulsive SE to chronic nonconvulsive SE recapitulates the clinical course of NORSE more closely than gradual epileptogenesis models. Together, these features fill a gap in available models for studying pediatric refractory epilepsies in general, and NORSE, specifically. Our hippocampally-focused model complements neocortical injury models of SE (Kharatishvili et al., 2006; Singh et al., 2022) that show the involvement of extra-hippocampal circuits implicated in psychiatric comorbidities in NORSE. Within the broader landscape of SE modeling, the hippocampal-focused IHKA model highlights complementary pathways to NORSE and the importance of examining diverse pathogenic pathways involved in NORSE.

An important consideration is whether the immune activation observed in our model reflects NORSE-specific pathology or simply represents inflammatory changes following prolonged seizures. Several lines of evidence suggest our model captures NORSE-relevant pathophysiology: (1) The specific immune cell populations infiltrating the hippocampus (CD8+ and γδ T cells) match those reported in NORSE patient autopsy studies (Dilena et al., 2019); (2) The temporal relationship between immune activation and seizure progression suggests immune-mediated mechanisms induced by prolonged and frequent repetitive seizures; (3) The age-dependent vulnerability (adolescent vs. adult) parallels the clinical epidemiology of NORSE. We acknowledge that dissecting primary immune-driven mechanisms from secondary seizure-induced neuroinflammation in NORSE pathophysiology requires additional studies, including immune depletion experiments, use of immunodeficient transgenic mice, and early-timepoint immunomodulation.

Neuroinflammation in the NORSE mouse model: key findings and therapeutic implications

Our findings of elevated microglial, astrocytic, and monocyte activation in this adolescent IHKA model underscore neuroinflammation as a central feature in disease pathology. In the absence of any proven disease-modifying therapy in NORSE, current expert consensus emphasizes early comprehensive evaluation, prompt initiation of immunotherapy, and consideration of ketogenic dietary interventions, although supporting evidence remains largely observational and limited (Wickstrom et al., 2022a). The clinical and economic burden of NORSE is substantial, with prolonged ICU stays and extended hospitalizations contributing to escalating healthcare costs (Wang et al., 2021; Schubert-Bast et al., 2022). In pediatric cases, the extended admissions impose an additional emotional and financial toll on families (Wang et al., 2021; Schubert-Bast et al., 2022). Even among survivors of children with NORSE, persistent multidomain neurocognitive impairments are common, often requiring specialized education and long-term care(Shrestha et al., 2023).

Recent preclinical and translational research implicates innate immune signaling pathways (e.g., IL-1β/IL-1R1 and HMGB1/TLR4) and broader neuroinflammatory cascades as critical mediators in seizure exacerbation and epileptogenesis (Choi and Koh, 2008; Vezzani et al., 2013; Zhang et al., 2022). However, these mechanistic targets cannot be ethically interrogated in humans, underscoring the need for translational animal models. The establishment of this adolescent IHKA mouse model, therefore, provides a platform to dissect the underlying immunopathology and identify novel therapeutic targets.

Histopathological studies on NORSE patients have documented marked reactive gliosis (Costello et al., 2009; van Baalen et al., 2010; Kramer et al., 2011; Howell et al., 2012), T cell-associated inflammatory changes (Dilena et al., 2019), and microglial and astrocytic activations (Costello et al., 2009), along with elevated proinflammatory cytokines (Choi and Koh, 2008). Our adolescent IHKA mouse model recapitulated these features. Specifically, adolescent (P40) IHKA mice exhibited a pronounced expression of astrocytes and microglia, a hallmark of neuroinflammation. Additionally, IHKA-Cx3Cr1 ^GFP/+ Ccr2 ^RFP/+ transgenic mice displayed significant monocyte infiltration in the brain, paralleling findings from previous SE models (Varvel et al., 2016). Similarly, increased numbers of CD8+ and γδ T cells were observed in the hippocampus of IHKA-treated adolescent mice, mirroring human NORSE biopsy findings. Together, these data suggest that sustained activation of inflammatory cells and immune pathways may contributes to seizure persistence and epileptogenesis in NORSE.

Furthermore, the blood-brain barrier (BBB), which regulates the entry of immune cells and molecules into the CNS, appears to be compromised in this model. Normally, the BBB restricts the entry of circulating immune cells and molecules, thereby preserving CNS homeostasis (Choi and Koh, 2008). While the brain’s intrinsic immune activation can facilitate repair from neurological insults (Correale and Villa, 2004), excessive infiltration of peripheral immune cells can exacerbate neuronal injury (Elenkov et al., 2005). Our data suggest that increased BBB permeability in IHKA-injected adolescent mice facilitates peripheral monocytes and proinflammatory molecules entry into the brain, which may exacerbate neuroinflammation, lower seizure threshold, enhance neuronal excitability, and contribute to epileptogenesis. This highlights the potential of targeting neuroinflammatory pathways, such as the interleukin-1 receptor/toll-like receptor (IL-1R/TLR) axis, as a promising therapeutic strategy for treating NORSE and related immune-mediated epilepsies.

The IHKA model has also been informative for studying the long-term cognitive and behavioral consequences of early-life seizures. Younger mice injected before P40 display enduring deficits in spatial memory and increased anxiety-like behaviors (Koyama et al., 2012; Levesque and Avoli, 2013). Consistent with these observations, early neurological and behavioral abnormalities emerged in our adolescent IHKA model within three weeks post-IHKA, consistent with the time line observed in othetr studies using younger animals. The IHKA model in younger mice has been instrumental in investigating neuroprotective strategies and anti-epileptogenic interventions, with some promising results in mitigating the long-term sequelae of early-life seizures (Tse et al., 2014; Umpierre et al., 2014). Our findings, alongside these studies, reinforce the utility of the IHKA paradigm at younger developmental stages for investigating age-dependent vulnerability and developing tailored interventions for pediatric epilepsy.

An important limitation of the current study is that we did not test the efficacy of conventional ASMs against the acute convulsive SE or chronic spontaneous seizures. A crucial next step to validate the model's utility for modeling treatment-refractory aspects of NORSE would be to systematically evaluate ASM responsiveness to determine whether seizures in this adolescent IHKA model exhibit the pharmacoresistance characteristic of human NORSE. Nevertheless, pharmacoresistance of status epilepticus has previously been shown in a variety of animal models of SE (Bankstahl and Loscher, 2008; Fritsch et al., 2010) following delayed treatment with benzodiazepines and phenobarbital. Moreover, resistance of post-IHKA seizures to anti-seizure medications (ASMs) has been well-established (Wang et al., 2023). Widmann et al., for example, demonstrated that focal seizures in the IHKA model did not respond to commonly used ASMs including diazepam, levetiracetam, lamotrigine, and oxcarbazepine (Widmann et al., 2024).

In summary, our adolescent IHKA model in P40 mice provides a robust experimental framework for elucidating the immunopathogenesis of NORSE and evaluating potential therapeutic strategies. The concurrent activation of both innate and adaptive immune pathways reveals a complex immune landscape underlying seizure persistence in this model. This work not only advances mechanistic understanding of NORSE but also identifies translational opportunities for developing targeted therapies for drug-resistant epilepsies.

Supplementary Material

Figure S1 and legend

Figure S1. Average electrographic seizure duration in the adolescent IHKA NORSE model.

Average duration of individual electrographic seizures at days 3, 5, 7, and 21 post-IHKA injection. Bars represent group, means ± SEM, with individual mouse data shown as overlaid circles (n=8 mice per timepoint). Average seizure duration increased significantly from Day 3 (13.2 ± 1.2 sec) to Days 5, 7, and 21 (19.5 ± 0.8, 18.6 ± 2.2, and 20.4 ± 1.5 sec, respectively). Analysis of 1,455 individual seizures revealed that 66% were brief (5–20 seconds), 15% were moderate duration (20–30 seconds), and 19% were prolonged (>30 seconds). Statistical analysis: one-way ANOVA with Tukey's post-hoc test; *p < 0.05.

Highlights.

New-onset refractory status epilepticus (NORSE) is a devastating neurological condition resistant to therapy with extremely high morbidity and mortality rates. NORSE often affects previously healthy children and young adults. A lack of clinically relevant animal models has hampered research progress and testing of new therapeutic approaches. We developed an adolescent mouse model that faithfully recapitulates the key features of NORSE, including subclinical status epilepticus accompanied by neuroinflammation, cell death, and cognitive impairment. This model showed marked microglial activation and robust immune cell infiltration in the hippocampus, providing critical insights into the immune mechanisms potentially contributing to NORSE pathogenesis. Our murine model can serve as a research platform to advance our understanding of NORSE and test anti-inflammatory, immunomodulatory therapies.

Data availability statement

The datasets generated for this study are available upon request to the corresponding author.