Sprague Dawley (SD) rats were housed in an animal enclosure with a temperature of 60% humidity, a temperature of 21 ± 1 °C, and a 12 h/12 h light/dark cycle. They were fed with food and water ad libitum. This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Committee on the Ethics of Animal Experiments of Chongqing Medical University (Chongqing, China). All surgery was performed under sodium pentobarbital anesthesia, and all efforts were made to minimize suffering.

SE was induced in 20-day-old male SD rats via intraperitoneal (i.p.) injection of pilocarpine (30 mg/kg, Sigma-Aldrich, St. Louis, MO, USA) 18–24 h after lithium chloride injection (127 mg/kg, Solarbio, Beijing, China). Each rat was randomly assigned to either the experimental group or the control group. Thirty minutes before pilocarpine, atropine (Aladdin Biochemical Technology, Shanghai, China) l mg/kg was intraperitoneally injected to reduce mortality. Rats in the experimental group were intraperitoneally injected with 10 mg/kg diazepam to terminate the seizures 60 min after SE, while rats in the control group were injected with an equal volume of sterile saline. Animals showing no seizures, subthreshold Racine grades (Racine scores <3), or seizure-related death were excluded. All behavioral seizure scoring was performed by two independent blinded observers, with discrepancies resolved by a third blinded evaluator. Seizure severity was assessed using the Racine scale, which classifies epileptic behaviors into 5 progressive stages based on motor manifestations (Racine et al., 1972). Score 1: Facial clonus (e.g., blinking, twitching of vibrissae, rhythmic jaw movements); Score 2: Head nodding accompanied by facial clonus; Score 3: Unilateral forelimb clonus with preserved posture; Score 4: Bilateral forelimb clonus and rearing (loss of balance); Score 5: Generalized tonic-clonic seizures with rearing, falling, and severe convulsions.

Treatment groups were randomly assigned using computer-generated block randomization (stratified by litter), with cages labeled by coded identifiers to conceal group identity. 20-day-old male SD rats in the experimental group were randomly assigned to the SE or N-[N-(3,5-difluorophenacetyl)-l-alanyl)]-S-phenylglycine t-butyl ester (DAPT) group. Specifically, 5 μl of DAPT (D5942, Sigma-Aldrich) dissolved in DMSO (50 μM) were injected stereotaxically (0.5 μl/min) into the lateral ventricle of rats in the DAPT treatment group using a Hamilton syringe at 30 min after seizure termination (ether inhalation anesthesia). Stereotaxic coordinates relative to bregma were: anteroposterior (AP): −0.8 mm, mediolateral (ML): ±1.5 mm, dorsoventral (DV): −3.5 mm (based on the rat brain atlas for adolescent rodents). The needle was left in place for an additional 5 min after injection to allow for diffusion of the drug, then slowly withdrawn over 2 min to prevent backflow. Conversely, rats in the SE group only received the same volume of DMSO as the vehicle. The experimental design is presented in Figure 1.

This experiment lasted for 2 years. A total of 110 rats were used for the experiment. Analyses were performed on the same cohort of animals to enable direct structure-function correlation, with some tissues from this cohort used for other assays in the previous study (Yuan et al., 2020). Animal health and behavior were monitored once everyday. Six rats died after electrode implantation. Twenty rats died during or after SE induction, and 20 rats died during electroencephalogram examination (Table 1). Rats were euthanized 28 days after SE induction via an intracardial perfusion of sodium sulfide-containing perfusate after sodium pentobarbital anesthesia. All animal welfare considerations were taken, including efforts to minimize suffering and distress, use of analgesics. All research staff got special training in animal care or handling. Histological processing and image analysis were conducted by technicians unaware of group assignments, with slides relabeled prior to quantification.

Electroencephalogram monitoring were performed according to the previously published method (Han et al., 2021). For anesthesia, we used a ketamine-xylazine mixture (75 mg/kg + 5 mg/kg, i.p.) with supplemental doses (1/3 initial dose every 30 min) adjusted for pup weight. A hole was drilled with a diameter of 1.00 mm on each side of the sagittal suture of the rat parietal bone (n = 6/group), and grounded it above the telencephalon. Then a self-made simple electrode (0.1 mm nickel chromium wire) was inserted and contacted the dura mater for scalp electroencephalogram examination. The electrode assembly was secured using a combination of medical-grade dental cement and biocompatible adhesive to create a stable interface that accommodates skull growth while maintaining electrode positioning. Post-operative analgesia consisted of buprenorphine (0.1 mg/kg, s.c.) administered every 8 h for 48 h. Pups were recovered on a 32 °C heating pad for 1 h before being transferred to foster dams (acclimated to pup scent), with daily monitoring of weight and nursing behavior. For pups with feeding difficulties, we provided 0.2 mL of sterile milk formula (20% skim milk + 5% glucose/electrolytes) via gastric tube every 4 h, maintaining separation for 48 h until preoperative weight was regained. After 4 days of wound healing, induce SE according to the above method. The inclusion criteria for SE induction required: ≥5% weight gain over 24 h, normal feeding/exploratory behaviors, intact tactile responses, and absence of neurological abnormalities (e.g., seizures, ataxia) within 72 h post-op, verified through daily weighing and behavior (Table 2). During the onset of SE, perform EEG examination. Afterwards, perform 2 h (from 9:00 AM to 11:00 AM) of EEG examination every day until 60 days after SE induction. Adjust ADL instrument amplifier parameters based on previously published articles. According to a previous study (Han et al., 2021; Cao et al., 2004), seizures were defined as discharges with the following characteristics: frequency>5Hz, amplitude>2 times baseline, and duration ≥ 10s. Spike frequency amplitude, and duration were calculated by segmenting visually identified ictal epochs from 3 spontaneous seizure sessions per animal (n = 6 per group), averaging values across 5 consecutive spike-wave discharges per seizure, then across all seizures per session, and finally across animals to generate group-level data. The pClampfit software was used to record the average number of spontaneous seizures per day, and analyze the duration and discharge frequency of each seizure. Continuous impedance measurements and periodic visual inspections of the electrode-tissue interface were implemented, with any recordings showing impedance fluctuations >20% being excluded from analysis. The ADJUST plugin was employed to implement an automated artifact detection algorithm, which identified and excluded motion-related signals.

Timm staining was performed using the FD Rapid TimmStain™ kit according to the manufacturer's instructions (#PK701, FD NeuroTechnology, Columbia, MD, USA). Specifically, rats were sacrificed 28 days after SE induction (n = 6/group) via an intracardial perfusion with 100 mL of 0.1 M sodium sulfide-containing perfusate, followed by 250 mL of 4% paraformaldehyde (PFA) in 0.1 M phosphate-buffered saline (PBS, pH 7.4). Brains were postfixed in 4% PFA for 24 h at 4 °C and transferred to 30% sucrose solution in 0.1 M PBS for 72 h at 4 °C. Cryoprotected brains were sectioned coronally at 40-μm intervals using a Leica CM1950 freezing microtome (Leica Biosystems, Nussloch, Germany) and mounted on positively charged slides, air-dried at room temperature for 24 h, and then stored in a light-protected box at −20 °C until staining. Under light-protected conditions, the slides were immersed in freshly prepared Timm's developer within a Coplin jar, followed by incubation in a precisely controlled 26 °C water bath for 90–150 min to allow optimal metal sulfide precipitation. Post-staining, the slides were thoroughly rinsed 3–5 times with running tap water (maintained under dark conditions to prevent photochemical artifacts), air-dried in a desiccator, and finally coverslipped with DPX neutral mounting resin for long-term preservation. Images were acquired using an Olympus BX53 microscope (Olympus, Tokyo, Japan) equipped with a 20 × objective lens. Image analysis was performed using ImageJ software (version 1.53k, National Institutes of Health, Bethesda, MD, USA). The analysis of MFS in CA3 region was performed according to the Gavazos MFS Classification Standard (0–5 scale) (Gold, 2012), where grade 0 indicates no sprouting, grade 1 represents occasional sprouting fibers in the supragranular layer, grade 2 shows a sparse but continuous band of fibers, grade 3 exhibits a dense but narrow band, grade 4 displays a dense and wide band, and grade 5 demonstrates extensive sprouting forming a dense plexus extending into the molecular layer1. The quantification was conducted by two blinded investigators who analyzed Timm-stained coronal sections of the dorsal hippocampus under light microscopy, with inter-rater reliability >90%. Only the dorsal hippocampus was analyzed in this study. The final MFS score for each animal represented the average of three consecutive sections within the dorsal hippocampal formation. In the DG region, density measurements were performed using ImageJ. The region of interest was selected in the inner molecular layer of the DG, and the integrated optical density (IOD)/area of Timm staining was calculated to reflect the sprouting intensity.

Neurons in the CA1 and CA3 regions of the dorsal hippocampus from rats in each group were analyzed by Nissl staining 28 days after SE induction (n = 6/group). Brain tissues were immersion-fixed in 4% paraformaldehyde (PFA) at 4 °C for 72 h. Samples were dehydrated through a graded ethanol series (70% for 10 min, 80% for 10 min, 95% for 10 min, and 100% for 15 min) followed by xylene clearing for 30 min, then embedded in high-grade paraffin wax for microtome sectioning. The paraffin sections of brain tissues after different treatments were routinely deparaffinized, hydrated through an alcohol gradient (soaked in 95%, 80%, and 70% alcohol for 1 min each). Following a 1-min rinse with distilled water, Nissl staining was performed by uniformly applying 100–200 μL of cresyl violet solution per brain section, allowing the stain to develop for approximately 3 min. Subsequently, the sections were gently rinsed under low-flow tap water for 2 min to remove excess dye. Rapid differentiation was achieved by briefly immersing the sections in 95% ethanol, followed by dehydration in absolute ethanol, clearing in xylene, and final mounting with neutral balsam. Six rats in each group were selected to generate brain tissue specimens, and five sheets of each specimen were selected for Nissl staining.

The brain slices on the cover glasses were imaged using a fluorescence microscope (IX53, Olympus) equipped with a 40 × objective lens. Positive neurons were quantified using ImageJ software (US National Institutes of Health, Bethesda, MD, USA). The analysis was performed using a semi-automated approach: images were first thresholded to identify neuronal somata, followed by manual verification to exclude non-specific staining artifacts. All analyses were conducted by an observer blinded to the experimental group assignments.

TEM was performed to observe the changes of synaptic morphology and density in the hippocampal CA1 region 28 days after SE induction in the different treatment groups (n = 6/group). Brain tissues were harvested immediately after transcardiac perfusion with 100 mL of 0.1 M sodium sulfide solution, followed by 250 mL of 4% paraformaldehyde (PFA) in 0.1 M phosphate-buffered saline (PBS). The perfused brains were post-fixed in the same 4% PFA solution at 4 °C for 24 h, then transferred to 30% sucrose in 0.1 M PBS for cryoprotection until they sank to the bottom. Hippocampal CA1 tissues were fixed in 2.5% glutaraldehyde for 2 h at 4 °C, washed three times with 0.1 M PBS, and stained with 2% OsO₄ for 2 h at room temperature. After dehydration via a series of alcohol concentrations (30%, 50%, 70%, 80%, 95%, and 100%) and acetone, the tissue was embedded in epoxy resin to form small 5-mm spheres. Cut the embedded tissue blocks into ultra-thin sections (60 nanometers thick) using a Leica EM UC7 ultramicrotome (Leica Microsystems, Wetzlar, Germany), and stain them with uranyl acetate and lead citrate. Observation was then performed by TEM. The criteria used to identify ultrastructural synaptic profiles were previously described (Gold, 2012; Szule et al., 2015; Han et al., 2022). In order to analyze the morphological changes in the length of the presynaptic membrane active band (AD), synaptic gap (SG) width, and postsynaptic density (PSD) thickness of the hippocampus at the ultrastructural level, 5 slices were randomly selected from each sample according to the principle of systematic random sampling and photographed using TEM (magnification, × 40,000). A total of 10 micrographs were captured for each rat. The AD length, SG width, and PSD thickness were measured by hand-tracing using ImageJ. For AD length, the freehand line tool was used to trace the full linear extent of the electron-dense presynaptic active zone and measurements were recorded in nanometers (nm); for SG width, a straight line tool was applied perpendicular to the synaptic cleft at three equidistant points along the AD and the mean value was calculated; for PSD thickness, the perpendicular distance from the postsynaptic membrane to the outermost edge of the electron-dense PSD region was measured at three points and the mean value was determined for each synapse.

All data were recorded as the mean ± SD and statistically analyzed using SPSS 17.0 software (IBM, Armonk, NY, USA). An unpaired Student's t-test was used for comparisons between two groups. For multiple comparisons, one-way analysis of variance with Dunnett's T3 test was used. All datasets were assessed for normality using Shapiro-Wilk tests (α = 0.05) and homogeneity of variance with Levene's test. Parametric tests (ANOVA/t-test) were applied only when both assumptions were satisfied; otherwise, Mann-Whitney U test for independent groups was used. p < 0.05 was considered statistically significant.

All relevant data generated in this research can be requested from the corresponding author.

Our system demonstrated stable recording capabilities throughout the 2-month period, as evidenced by consistent electrode-skin interfacial impedance values (maintained within 15% of baseline) and reproducible signal characteristics across recording sessions. The inhibitory effect of DAPT on the Notch signaling pathway has been experimentally validated through Western blot and immunofluorescence assays, as demonstrated in the authors' previously published work (Yuan et al., 2020).

Epileptogenesis is a pathological process in which sporadic recurrent seizures (SRSs) gradually appear after SE. Rats in both the SE and DAPT groups could detect varying degrees of SRS during the chronic phase of epileptogenesis afer SE, manifested as scattered spike and spike waves lasting for tens of seconds (Figure 2a). Frequent epileptiform discharges occurred in the acute phase (1–3 days) after SE induction. The number of epileptiform discharges increased slightly over time, but the duration of each epileptiform discharge shortened. By contrast, although SRSs also appeared in the chronic period after SE induction in the DAPT group, the amplitude of the sharp and spiky waves decreased, and the frequency of the discharges slowed. In DAPT group, the average number of spontaneous seizures per day [12.55 ± 1.12 vs. 8.29 ± 0.86, p = 0.019, 95%CI (0.77, 7.76), Figure 2b], duration of seizure [43.83 ± 2.24 vs. 35.77 ± 2.59, p = 0.038, 95%CI (0.49, 15.62), Figure 2c] and discharge frequency of each seizure [9.94 ± 0.93 vs. 5.73 ± 0.65, p = 0.006, 95%CI (1.36, 7.07), Figure 2d] were significantly lower compared with that in SE group. Moreover, the degree of epileptic seizure behavior in the DAPT group rats was significantly reduced compared to the SE group (Table 2). In the DAPT group, seizures in rats were mainly characterized by binocular gaze and facial muscle twitching. In the SE group, rats typically exhibit seizure behavior with unilateral or bilateral forelimb spasms. No statistically significant difference was observed in the latency period from acute SE to the onset of SRS between the SE group and the DAPT-treated group, indicating that DAPT intervention did not affect the latency to spontaneous epileptic seizures. These results suggest that Notch signaling blockade after SE induction partially attenuates the severity of EEG epileptic discharges.

The presynaptic membrane active band is the specialized site for neurotransmitter release where synaptic vesicles dock and fuse. Its structural integrity is critical for regulating vesicle release probability and location, which directly impacts synaptic transmission reliability and neural coding (Rozenfeld et al., 2023). The SG width influences the spatiotemporal dynamics of neurotransmitter diffusion and receptor activation. Pathological cleft expansion may decouple pre- and postsynaptic signaling (Jiang et al., 2024). The PSD contains scaffold proteins that organize glutamate receptors and signaling complexes. Its thickness reflects the density of these molecular machines, which determines synaptic strength and plasticity (Heo et al., 2023). Reduced PSD thickness correlates with diminished excitatory postsynaptic currents and cognitive deficits. The effects of Notch signaling blockade on the synaptic ultrastructure of the CA1 region after SE induction were investigated by TEM (Figure 3a). The AD length, SG width, and PSD thickness were measured in each group. The results demonstrated that in the SE group the AD length was significantly reduced [238.6 ± 8.14 vs. 185.5 ± 8.44, p = 0.0011, 95%CI [(6.99, 79.24), Figure 3b], the SG was widened [18.32 ± 1.21 vs. 22.78 ± 0.83, p = 0.0121, 95%CI (−7.71, −1.21), Figure 3c], and the PSD was blurred compared with the control group. In addition, the thickness of the PSD was reduced [64.08 ± 1.15 vs. 53.10 ± 3.05, p = 0.0091, 95%CI (3.39, 18.58), Figure 3d]. By contrast, DAPT administration partially reversed the reductions of the AD length [185.5 ± 8.44 vs. 212.9±6.87, p = 0.0306, 95%CI (−51.62, −3.14), Figure 3b]and PSD thickness [53.10 ± 3.05 vs. 61.67 ± 1.51, p = 0.0306, 95%CI (−16.16, −0.98), Figure 3d], but it had no significant effect on the SG width (Figure 3c). This suggests that inhibiting the activation of Notch signaling after SE induction can promote the repair of the synaptic ultrastructure.

The structure and damage of hippocampal CA3 and CA1 pyramidal neurons in each group were examined by Nissl staining 28 days after SE induction. The results showed that the arrangement of hippocampal CA3 pyramidal neurons in the SE group was loose, and some cells exhibited wrinkles or vacuolization (Figure 4a). Notably, compared with the control group findings, there were fewer Nissl bodies in the pyramidal cells of the CA3 [32.50 ± 4.85 vs. 83.78 ± 4.92, p < 0.0001, 95%CI (−66.60, −39.95), Figure 4b] and CA1 [27.90 ± 4.62 vs. 62.40 ± 7.70, p = 0.0033, 95%CI (−54.51, −14.48), Figure 4c] regions of the hippocampus in the SE group, and these findings were accompanied by scattered nucleolysis and disorganized cell arrangement. After DAPT administration, the disorganization of neuronal cells in the CA1 and CA3 regions was significantly improved, and the cellular hierarchy was distinct. The cell size and morphology were basically normalized (Figure 4a), and the number of homogeneously stained Nissl granules in the cytoplasm was increased in the CA3 (57.99 ± 5.60 vs. 32.50 ± 4.80, p = 0.0062, 95%CI (15.62, 35.36), Figure 4b] and CA1 (49.19 ± 5.85 vs. 27.90 ± 4.62, p = 0.017, 95%CI (12.4, 30.2), Figure 4c] regions compared with SE groups. This suggests that after SE induction, the inhibition of Notch signaling by DAPT partially reduced the occurrence of neuronal loss in hippocampal CA3 and CA1 region.

Timm staining was applied to observe changes in the distribution and number of Timm-stained granules to characterize the distribution of MFS and synaptogenesis. Timm scores and MFS densities were examined on the slices of hippocampal regions from different groups. The Timm scores in the hippocampal CA3 region was higher than that in the control group [3.61 ± 0.35 vs. 1.10 ± 0.28, p = 0.0002, 95%CI (1.52, 3.49), Figure 5b], which also exhibited a significantly higher number of Timm-stained particles (Figure 5a) and MFS density in the DG region was significantly increased in the SE group [49.19 ± 5.85 vs. 27.90 ± 4.62, p = 0.017, 95%CI (12.4, 30.2), Figure 5c], In the DAPT group, both the Timm scores in the CA3 region and MFS densities in the DG region were significantly increased compared to the control group, but no significant difference was observed when compared to the SE group, indicating that DAPT treatment failed to improve MFS after SE induction.