Exploratory analysis of a Novel RACK1 mutation and its potential role in epileptic seizures via Microglia activation

Abstract

Seizures is a prevalent neurological disorder with a largely elusive pathogenesis. In this study, we identified the key gene RACK1 and its novel mutation RACK1-p.L206P as being associated with seizures through single-cell transcriptome sequencing (scRNA-seq) and whole exome sequencing (WES) techniques. Our findings reveal that the RACK1-p.L206P mutation significantly enhances proliferation, migration, phagocytic ability, and inflammatory activation in human microglia, which in turn affects neuronal excitability and synaptic function, culminating in typical seizure symptoms in the seizures. These effects were further validated in a mouse model using CRISPR/Cas9 gene editing technology. Mutant microglia exhibited increased activation and induced apoptosis in hippocampal neurons, leading to higher action potential frequency and excitatory synaptic marker expression. In vivo experiments demonstrated that RACK1-p.L206P mutant mice displayed classic seizure symptoms, with increased neuronal excitability and a tendency for action potential bursts during initial depolarization, along with more frequent spike discharges. Additionally, excitatory synapse density and size in the hippocampal CA1 region of mutant mice were significantly elevated, accompanied by increased expression of VGLUT1 and PSD95 within microglia. This study offers novel insights into the molecular mechanisms underlying seizures in the seizures and presents valuable clues for the development of future therapeutic strategies.

Introduction

Seizures are neurological events triggered by abnormal electrical discharges in the brain, resulting in symptoms such as convulsions, altered consciousness, and other physiological responses [1]. seizures commonly occur in patients with epilepsy - a chronic disorder characterized by recurrent, unprovoked episodes [2]. However, seizures can also occur as isolated incidents, triggered by acute brain injury, infection, or induced states such as status epilepticus [3, 4]. Understanding the mechanisms behind seizure occurrence, whether in the context of chronic epilepsy or as isolated events, remains a crucial area of neurological research. In recent years, rapid advances in molecular biology and genomics have greatly enhanced our understanding of seizure mechanisms, especially through technologies such as Whole Exome Sequencing (WES) and scRNA-seq. These tools have illuminated the genetic and molecular underpinnings of seizure susceptibility and response, paving the way for novel therapeutic strategies to alleviate the burden of seizures on patients and their families.

In recent years, with the rapid progress in technologies such as molecular biology, genomics, and transcriptomics, there has been significant advancement in our understanding of the genetic background and molecular mechanisms of seizures. The application of techniques such as wes and scRNA-seq has provided new perspectives for exploring the genetic characteristics and molecular mechanisms of seizures [5]. Previous studies have revealed many gene mutations and pathway alterations related to seizures, but their specific roles and molecular mechanisms require further investigation [6,7,8].

Recently, significant progress has been made in the application of CRISPR/Cas9 gene editing technology in neuroscience. For instance, researchers have successfully employed this technology to precisely modify key genes associated with neurodegenerative diseases and psychiatric disorders in various animal models, elucidating the functions and disease mechanisms of these genes [9, 10]. These groundbreaking studies not only enhance our understanding of diseases but also provide an experimental platform for developing new therapeutic strategies.

The pathophysiological mechanisms of seizures are complex, involving disrupted neural networks and neurotransmitter imbalances, necessitating a multifaceted research approach [11, 12]. Microglia play a key role in neural immune regulation and inflammatory responses, contributing to the pathogenesis of various neurological disorders, including seizures [13,14,15]. RACK1, a multifunctional protein, has unclear roles in microglia [16]. The specific functions and mechanisms of RACK1 in seizures remain poorly understood. While the role of the RACK1 gene in seizures has not been extensively investigated, preliminary evidence indicates that RACK1 may modulate the interaction between neurons and microglia, thereby affecting neuroimmune responses and cell death pathways [17]. Furthermore, certain known mutations in RACK1 have been closely associated with the development of other neurological disorders, warranting further exploration of its specific role in seizure onset.

In this study, we employed a combination of scRNA-seq and WES technologies to reveal the genetic characteristics and pathogenic mechanisms of seizures. We particularly focused on a newly discovered RACK1, RACK1-p.L206P mutation, and how it triggers seizures by altering microglial function. By delving into the role of RACK1 mutations in seizure onset, we aim to provide new theoretical foundations and practical directions for the diagnosis, treatment, and prevention of seizures. The findings of this study may contribute to the development of novel therapeutic strategies for seizures, offering clinicians more targeted treatment options to improve the prognosis and quality of life for seizure patients.

Materials and methods

Animal modeling of status epilepticus model

Seventy healthy C57BL/6J mice (60 males, 10 females), aged 6–8 weeks, were obtained from our animal experimental center. Each mouse was individually housed in an SPF-grade facility, with controlled conditions of 60−65% humidity and a temperature range of 22–25 °C, under a 12-hour light-dark cycle. Food and water were provided ad libitum. After a one-week acclimation period, experiments were initiated. All procedures were approved by the institutional animal ethics committee.

Anesthesia was induced with chloral hydrate (6001-64-5, Sigma Aldrich, USA), after which the heads of 8–10-week-old male mice were secured in a stereotaxic apparatus. The fur along the midline was shaved, and the area was sterilized. A small incision was made to expose the skull. Using stereotaxic coordinates, the hippocampal region was accurately located. A 50 nL injection of either 20 mM kainic acid (KA) (487-79-6, Sigma Aldrich, USA) or 0.9% NaCl (7647-14-5, Sigma Aldrich, USA) was administered into the basolateral nucleus of the right amygdala. After the injection, the incision was sutured for proper healing. Mice were allowed to recover from anesthesia for approximately 2 h and subsequently monitored for 8 h. Seizure severity was evaluated using the Racine scale, which classifies stages as follows: Stage 1, facial clonus; Stage 2, head and neck rigidity; Stage 3, forelimb clonus; Stage 4, generalized clonic-tonic seizures; and Stage 5, generalized tonic-clonic seizures with loss of posture and convulsions. Status epilepticus was defined by the occurrence of Stage 4 or 5 seizures [18].

Following KA administration, three monopolar electrodes were implanted in the bilateral frontal cortex and cerebellum (reference electrode). EEG recordings of freely moving mice were captured using a digital acquisition system (Coherence, Deltamed, France; sampling rate 256 Hz) starting immediately upon recovery from anesthesia. Observations indicated no seizure-like activity in the cortex within the initial 6 h post-KA injection, although hippocampal spikes and multi-spikes were noted. At 12 h, sharp waves were detected in both the hippocampus and ipsilateral cortex, with occasional sharp wave events recorded intermittently in either the cortex or hippocampus between 22 and 24 h (Figure S1). To maintain objectivity and accuracy, a double-blind design was employed, with data collection and analysis performed by a third party unaware of the experimental conditions, minimizing potential biases [19].

scRNA-seq and bulk RNA-seq

Following the onset of seizures in mice, the mice were euthanized, and hippocampal tissues from successfully modeled epileptic and normal mice (3 males per group) were collected. The tissues were dissociated into single-cell suspensions using trypsin (Sigma Aldrich, USA, 9002-07-7) and individual cells were captured using the C1 Single-Cell Auto Prep System (Fluidigm, Inc., South San Francisco, CA, USA). After cell capture, cells were lysed and mRNA was released within the chip, followed by reverse transcription to generate cDNA. The post-lysis and reverse transcription cDNAs were pre-amplified in a microfluidic chip for subsequent sequencing. The amplified cDNAs were then used for library construction and single-cell sequencing on the HiSeq 4000 Illumina platform (parameters: paired-end reads, read length 2 × 75 bp, approximately 20,000 reads per cell) [20]. Total RNA of at least 1 µg was isolated from hippocampal tissues of epileptic and normal mice, treated with DNAse I, and purified using silica membrane columns (Qiagen, Hilden, Germany, RNeasy Kit, 74004). RNA quantification was performed using Qubit RNA HS Assay Kit (Themo Fisher, USA, Q32852) and diluted to 100 ng/µL. The quality of total RNA was confirmed using a fragment analyzer (Agilent 5400, USA).

Samples with RNA quality and quantity ≥ 6 underwent RNA library preparation accredited by ISO/IEC-17,025 (Illumina, USA, TruSeq RNA Library Prep Kit v2, RS-122–2302). mRNA was enriched and fragmented using oligo dT magnetic beads (Suzhou Vazyme Biotech Co., Ltd., Suzhou, China, MS04T), followed by cDNA synthesis, adapter ligation, and PCR amplification. Paired-end sequencing with a read length of 126 bp was performed on the Illumina HiSeq 2500V4 sequencer according to the manufacturer’s protocol, generating at least 12.5 Gbp per sample, approximately 50 million read pairs. Image analysis, base calling, and quality control were conducted using Illumina data analysis pipelines RTA v1.18.64 and Bcl2fastq v1.8.4. RNAseq reads were provided in a compressed Sanger FASTQ format [21].

scRNA-seq data analysis

The data was analyzed using the “Seurat” package in R software. Data quality control was conducted based on the criteria of 200 < nFeature_RNA < 5000 and percent.mt < 20, leading to the selection of the top 2000 highly variable genes by variance [22].

To reduce the dimensionality of the scRNA-seq dataset, principal component analysis (PCA) was performed using the top 2000 highly variable genes. The Elbowplot function of the Seurat package was utilized to select the top 20 principal components (PCs) for downstream analysis.

The FindClusters function provided by Seurat was employed to identify major cell subpopulations, with the resolution set at the default value of 1. Subsequently, the t-SNE algorithm was applied to reduce the nonlinear dimensionality of the scRNA-seq sequencing data. Markers specific to various cell subpopulations were chosen using the Seurat package, and then cell annotation was conducted using the “SingleR” package [23]. Cell-cell communication analysis was carried out using the “CellChat” package in R language, and cell trajectory analysis was performed using the “Monocle2” package.

Differentially expressed genes (DEGs) screening and GO and KEGG enrichment analysis

DEGs in scRNA-seq datasets were identified using the “Limma” package in the R software. DEGs between Macrophages and Microglia in the Hipp of normal mice and seizure mice were filtered based on the criteria of |logFC| > 0.5 and P.adjust < 0.05 [24]. Moreover, the “Limma” package in R software was applied to screen for DEGs in RNA-Seq datasets, setting the standards at |logFC| > 1 and P.adjust < 0.05 for DEG selection between normal mice and seizure mice in the Hipp. Subsequently, significant DEGs underwent Gene Ontology (GO) functional and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses using the clusterProfiler package in R software, followed by data visualization analysis with the ggplot2 package [25].

Clinical samples

Our study enrolled 192 patients diagnosed with seizures at our hospital between January 2021 and December 2022. Peripheral blood samples were collected from these patients, and DNA was extracted within two weeks using the QIAamp DNA kit (56304, Qiagen, USA). The extracted DNA samples were stored at − 80 °C. All patients provided informed consent and the study received approval from the ethics committee, adhering strictly to the Helsinki Declaration. The age range of these patients was 30–75 years, with a mean age of 55.8 ± 8.2, comprising 121 males and 71 females. The diagnostic criteria for seizures were based on the guidelines revised in 2017 by the International League Against Epilepsy (ILAE) [26, 27].

Inclusion criteria: (1) This study enrolled adult patients aged between 30 and 75 years. All patients were diagnosed with seizures at our hospital between January 2021 and December 2022. (2) The patients included two main types of seizures, focal seizures where seizures originate from specific brain regions, and generalized seizures where seizures affect multiple brain regions.

Exclusion Criteria: (1) Other Major Diseases: Patients with hypertension, diabetes, hyperlipidemia, malignant tumors, severe cardiovascular or cerebrovascular diseases, and those requiring major organ transplants were excluded. (2) Severe Non-Neurological Conditions: Exclusion criteria included chronic obstructive pulmonary disease, chronic gastritis, gastric ulcers, duodenal ulcers, and similar conditions outside the neurological system. (3) Severe Neurological Pathologies: Patients with severe brain injuries, severe Parkinson’s disease, and severe mental disorders were excluded.

WES

We extracted genomic DNA from 192 seizure patients’ peripheral blood using the QIAamp DNA Kit (56304, Qiagen, USA). Exome capture, high-throughput sequencing, and general screening services were provided by Novogen Bioinformatics Institute in Beijing, China. All exons were captured using the Agilent SureSelect Human All Exon V6 Kit (43513, Agilent, USA) and sequenced on the Illumina HiSeq2000 platform. The screening strategy was in accordance with our previous study [28].

Prediction of pathogenicity and conservation of mutation sites

The impact of a non-synonymous mutation (c.617T > C; p. Leu206Pro) identified through WES on the protein was analyzed using the PolyPhen-2 database (http://genetics.bwh.harvard.edu/pph2/index.shtml). By inputting the UniProt Accession Number of the gene and the amino acid mutation (p. Leu206Pro) on the website, a score between 0 and 1 is generated, where a score closer to 1 indicates a higher level of pathogenicity [29].

The Mutation Taster database (https://www.mutationtaster.org/) was utilized to assess the harmfulness of the mutation. Upon entering the NCBI Gene ID and the mutation site (c.617T > C) along with the six nucleotides before and after, the output indicated the mutation as “disease causing” with predictions of phylop and phastcons conservation scores [30].

By subjecting the nucleotide sequence (including the mutation site along with the ten nucleotides before and after) to BLAST in the UCSC database (http://genome.ucsc.edu/), the conservation of the corresponding amino acid sequence in nine species was identified. The “ggseqlogo” package in R language was employed to analyze conservation.

The gnomAD database (https://gnomad.broadinstitute.org/), the world’s largest public DNA genetic mutation database, was consulted to determine whether the mutation in this study was novel.

Cell culture

Microglia (HMC3) (SNL-320, Sino Biological, Wuhan, China) and HPPNCs (BFN60803979, Green Quark Bio, Shanghai, China) were cultured using DMEM medium (11965092, Gibco, USA) containing 10% FBS, 10 µg/mL streptomycin, and 100 U/mL penicillin. The cells were cultured at 37 ℃ with 5% CO₂ in a humidified incubator (Heracell™ Vios 160i CR CO₂ incubator, Thermo Scientific™, Germany). Passaging was performed when cell confluency reached 80–90% [31,32,33].

Microglia activation by LPS: Microglia activation was induced by adding LPS (100 ng/mL) (L5293, Sigma Aldrich, USA) to HMC3, followed by a 24-hour stimulation for subsequent experiments [34].

For gene transfection, the pCAGGs-GFP plasmid (HG-VPH1278, Equl, Shanghai, China) carrying green fluorescent protein (GFP) was introduced into HPPNCs using lipofectamine 2000 (11668030, Thermo Fisher, USA). After 24 h, co-cultivation with HMC3 cells of different genotypes was conducted [35].

To investigate the impact of microglia on neurons, the supernatant collected from the culture medium 24 h after LPS stimulation of HMC3 was mixed in a 1:1 ratio with regular culture medium and added to HPPNCs for a 24-hour incubation before subsequent experimental analysis [34].

CRISPR/Cas9 gene editing technology

The RACK1-deficient cells were constructed using CRISPR/Cas9 technology with the following sgRNA sequences: RACK1-sgRNA: Forward: 5′-GGATCCCTCTGTGCTTCTGG-3′ (PAM: AGG), Reverse: 5′-GCCTGTGTGGCCAATGTGGT-3′ (PAM: TGG). The sgRNA was inserted into the Lenti-CRISPR v2 vector containing the Streptococcus pyogenes Cas9 nuclease gene (Hanheng Biotech, Shanghai, China). Cells were transduced with the lentiviral Lenti-CRISPR v2 vector and the CRISPR/Cas9 editing system was employed to generate the RACK1-L206P (c.617T > C) mutation, which was introduced into HMC3 cells (constructed and provided by EdiGene, Guangzhou, China). The donor sequence is as follows: GTATGGAACCTGGCTAACTGCAAGCTGAAGACCAACCACATTGGCCACACAGGCTATCTGAACACGGTGACTGTCTCTCCAGATGGATCCCCCTGTGCTTCTGGAGGCAAG.

Transfected cells with sgRNA plasmid and donor sequence were selected with 4 µg/mL puromycin. Surviving cells were clonally expanded and RACK1-L206P heterozygous and homozygous mutant cells were obtained through DNA sequencing. PCR and Sanger sequencing were performed: DNA was extracted from each clone, the target region of the RACK1 gene was amplified using specific primers, and Sanger sequencing was conducted to confirm whether the c.617T > C mutation guided by sgRNA was generated as expected. Confirmation of heterozygous and homozygous mutations: By comparing the proportions of wild-type and mutant sequences in the sequence chromatograms, heterozygous and homozygous mutant cells were distinguished. Homozygous mutant cells exhibited mutant peaks on both DNA strands, whereas heterozygous mutant cells showed the coexistence of wild-type and mutant peaks [36, 37].

Construction of RACK1-L206P Mutant Mice:

Construction of mouse-specific RACK1-sgRNA: 5′-AAGCTAAAGACCAACCACAT-3′ (Forward direction; PAM: TGG).

Mouse-specific Donor sequence:

GTGTGGAATCTGGCTAACTGCAAGCTAAAGACCAACCACATTGGCCACACTGGCTACCTGAACACAGTGACTGTCTCTCCAGATGGATCCCCCTGTGCTTCTGGAGGCAAG

Hormone injections (hCG (9002-61-3, Sigma-Aldrich, USA)) were administered to female C57BL/6 mice aged 4–6 weeks. One day before microinjection, female mice were mated with males to obtain fertilized eggs. On the injection day, superovulated female mice were dissected to collect fertilized eggs by removing the ovaries and fallopian tubes. The cumulus cells were removed by digestion with hyaluronidase (37326-33-3, Sigma-Aldrich, USA), and the embryos were placed in drops of embryo culture medium (KSOM embryo culture medium (with amino acids), M1430, Nanjing Aibeilife Biotechnology Co., Ltd., Nanjing, China). Under a microscope, a smooth glass pipette was used to hold the fertilized egg, and Cas9 mRNA, sgRNA mRNA, and donor solution were directly injected into the male pronucleus of mouse-fertilized eggs using another fine glass needle. The mouse embryos were transferred to surrogate mothers for pregnancy. F0 pups were sequenced to select positive mice, then crossed with wild-type C57BL/6J mice. Sequencing was performed on the F1 generation (Primer sequences: Forward: 5′-CTGCAAGCTAAAGACCAACCA-3, Reverse: 5′-AAATACCTTGCCTCCAGAAGC-3′), ultimately identifying homozygous and heterozygous mutant mice [37].

Sanger sequencing

The RACK1 gene exons were amplified using Polymerase Chain Reaction (PCR). The PCR products were purified using the DNA purification kit (D7001, Zymo Research, USA) and directly Sanger sequenced on capillary electrophoresis using the Applied Biosystems 3730 DNA analyzer (4331247, Applied Biosystems™) [38, 39].

Western blot

Total proteins were extracted from HMC3, HPPNCs, and mouse Hippocampal tissues. Cultured cells and tissues were digested using trypsin (T4799-5G, Sigma-Aldrich, USA) and collected. Enhanced RIPA lysis buffer (AR0108) containing protease inhibitors provided by Wuhan Dr. De Co., Ltd. (Wuhan, China) was used for cell lysis. Protein concentration was measured using the BCA protein quantification kit (AR1189, Wuhan Dr. De Co., Ltd., Wuhan, China). Proteins were separated by SDS-PAGE, transferred onto PVDF membranes, and blocked with 5% BSA (9048-46-8, Solarbio, Beijing, China) for 1 h at room temperature. The membranes were then incubated overnight at 4 °C with the diluted primary antibodies (specific information in Table S1). After washing the membranes three times with PBST (5 min each time), Anti-Mouse-HRP secondary antibody (Cat # 7076, 1/5000; CST, USA) or Anti-Rabbit-HRP secondary antibody (Cat # 7074, 1/5000; CST, USA) was added and incubated for 1 h at room temperature. Membranes were washed with PBST three times (5 min each time). PBST was removed, and an appropriate amount of ECL working solution (Omt-01, Beijing Aomijia De Pharmaceutical Technology Co., Ltd., Beijing, China) was added for incubation at room temperature for 1 min. The excess ECL reagent was removed, and the membrane was sealed in plastic wrap, placed in a dark box, and exposed on X-ray film for 5–10 min for visualization and fixing. Image J analysis software was used to quantify the grayscale of the bands in the Western blot images, with GAPDH serving as the internal control [40, 41].

CCK-8 experiment

The logarithmic growth phase HMC3 cells were seeded at a density of 5 × 10³ cells per well in a 96-well plate. Subsequently, 10 µL of CCK-8 reagent solution from Shanghai BiyunTian Biotechnology Co., Ltd. (Shanghai, China) (C0038) was added to each well, and the plate was then incubated in a humidified 37 °C cell culture incubator. After 1 h, the absorbance at 450 nm for each well was measured using a Microplate Reader (abx700005) from Beijing Qiwei Yicheng Technology Co., Ltd [42].

MTT Assay for cell viability

Cells were seeded at a density of 1 × 10⁴ cells per well in a 96-well culture plate and were incubated for 24 h. Cells from different treatment groups were then incubated with 0.5 mg/mL MTT (298-93-1, Sigma Aldrich, USA) in a culture medium for 4 hours, and cell viability was determined by measuring the absorbance at 570 nm [43].

Enzyme-linked immunosorbent assay (ELISA)

Cell culture supernatant or suspension samples from mouse Hippocampal tissue were used with IL-1β ELISA kits (Human, ab214025; Mouse, ab197742), TNF-α ELISA kits (Human, ab181421; Mouse, ab208348), and IL-6 ELISA kits (Human, ab178013; Mouse, ab222503) from Abcam (UK). The antigens were diluted to the appropriate concentrations in coating buffer and incubated with 5% bovine serum at 37 °C to block the enzyme-labeled reaction wells for 40 min. The diluted samples were then added to the reaction wells, followed by enzyme-labeled antibodies and substrate solution. Finally, the reaction was terminated by adding 50 µL of stop solution to each well within 20 min, and absorbance was read at 450 nm on a plate reader (Bio-Rad, USA) to generate a standard curve for data analysis [44].

Nitrite determination

Total nitrate levels were determined using Griess reagent (G7921, Invitrogen, Thermo Fisher, USA). Following a 10-minute incubation at 37 °C, absorbance was read at 540 nm, and the total nitrate concentration (µM) was calculated from a standard curve [45].

Scratch migration assay

HMC3 cells of different genotypes were cultured in a 24-well plate. Scratches were made in the middle of each well using a sterile 200 µL pipette tip, followed by a PBS wash to remove any damaged cell debris. At various time intervals (6 h, 12 h, and 24 h), images of the cells migrating into the scratched area were captured, and the percentage of area occupied by the migrated cells was measured using Image J software [46].

Phagocytosis assay

To assess the phagocytic activity of different genotypes of microglia, HMC3 cells were seeded at a density of 1 × 10⁴ cells per well in a 96-well plate and then co-incubated with 0.125 mg/mL pHrodo™ Green E. coli (P35366, Invitrogen™, Thermo Fisher, USA) for 6 h. The quantification of phagocytic particles was done by measuring the fluorescence excitation/emission at 509/533 nm. Moreover, cells with engulfed fluorescent bio-particles were imaged using fluorescence microscopy [47].

Automated image acquisition

Utilizing transient transfection of fluorescently labeled proteins, we could longitudinally track individual neurons in culture for extended periods. Through the Zeiss Observer Z1 microscope (Germany), automated longitudinal tracking of neurons post-GFP transfection was conducted to study neuron survival rates. Images were automatically captured using a 10x long-distance objective, allowing the software to sequentially perform tasks like locating specific neuron regions, automatic focusing, image acquisition, and moving on to the next non-overlapping neuron area. This facilitated rapid and efficient scanning of multiple neuron regions in each well. Following complete image acquisition, the plate was returned to the incubator until the next scan. For typical survival experiments, 10 positions were used per well, with 4 wells per condition. Positions were randomly selected to ensure unbiased neuron sampling. A template with the same initial spatial locations was employed in the experiments to track the same neuron region [48].

Immunohistochemistry and Immunofluorescence

For immunohistochemistry, tissues or cells to be tested are fixed, embedded, sectioned, and dewaxed. Dewaxing removes the wax from the section, making it hydrophilic for subsequent immunostaining. The dewaxed tissue sections are treated with a specific Cleaved-CASP3 antibody (Human, SAB1305630, 1:50; Sigma Aldrich, USA) followed by an Anti-Rabbit-HRP secondary antibody (12–348, 1:1000; Sigma Aldrich, USA). DAB chromogen (ab64238, Abcam, USA) is used to visualize the binding sites of the secondary antibody to the primary antibody. The stained tissue sections are then dewaxed and coverslipped for observation under a microscope to record the expression of Cleaved-CASP3 [49].

Cells are washed with cold PBS and fixed with 4% paraformaldehyde (P885233, Macklin, Shanghai, China) for 15–30 min. Subsequently, the cells are permeabilized with 0.1% Triton (L885651, Macklin, Shanghai, China) for 15 min to penetrate the cell membrane. After two washes with PBS, the cells are incubated with 15% FBS in PBS at 5 °C overnight for 4 min.

Mouse brains are extracted after decapitation. The brains are fixed with 4% paraformaldehyde at 4 °C for 4 days, followed by cryoprotection in a 30% sucrose solution (57-50-1, Sigma Aldrich, USA) at 4 °C for two days until equilibrium is reached, and the brain sinks. The brain is frozen in liquid nitrogen and stored at − 20 °C until slicing. Slices with a thickness of 30 μm are obtained at − 22 °C using a cryostat, stored in a 24-well plate with cryoprotectant (20% distilled water, 20% PB2x containing 5.73% Na₂HPO₄ (7558-79-4, Sigma Aldrich, USA) and 0.624% NaH₂PO₄ (10049-21-5, Sigma Aldrich, USA), 30% ethylene glycol (107-21-1, Sigma Aldrich, USA), and 30% glycerol (56–81-5, Sigma Aldrich, USA). Free-floating Hippocampal slices are permeabilized with 1% Triton X-100 in PBS for 1 h, followed by blocking with 5% bovine serum albumin and 1% Triton X-100 in PBS.

Cells or tissues are incubated with antibodies against Iba-1 (Human and Mouse, ab178846, Abcam, USA; 1:100), VGLUT2 (Human and Mouse, PA5-110380, Thermo Fisher, USA; 1:100), PSD95 (Human and Mouse, MA1-046, Thermo Fisher, USA; 1:100), and RACK1 (Human and Mouse, PA5-17484, Thermo Fisher, USA; 1:100) at 4 °C overnight, followed by incubation with Cy3 or FITC-conjugated secondary antibodies. Observation uses a fluorescence microscope (Zeiss Observer Z1, Germany) [50].

To quantify the percentage of phagocytic microglia cells, a region of interest (ROI) containing the Dentate Gyrus (DG) is created for each captured image. A binary image is created by establishing the same threshold for Iba-1 fluorescence on each image, followed by erosion, dilation, and watershed processes for particle analysis. This results in an ROI library for each microglia. Phagocytosis is defined as the formation of enclosed three-dimensional pockets around apoptotic cell processes by microglia. Apoptotic cells are defined based on Hoechst staining as cells with chromatin structure (true chromatin and heterochromatin) loss, condensation, and/or fragmentation (pyknosis/karyorrhexis). For surface IMARIS reconstruction, images are taken with a Zeiss LSM 2020 DUO confocal laser scanning microscope at 63x magnification with 0.64 μm (1024 × 1024) Z-stacks, which are then opened in IMARIS (v.9.6, Oxford Instruments, UK) in their original format. Z-stacks are automatically reconstructed into 3D models without further image preprocessing [51].

IL-1β/Iba-1 co-localization follows the incubation of tissues with Iba-1 antibody (Human and Mouse, ab178846, Abcam, USA; 1:100) at 4 °C overnight, followed by washing with PBS to remove unbound primary antibody or dye. The tissues are then incubated with the anti-IL-1β antibody (Mouse, AB1413-I-AF488, 1:100, Sigma Aldrich, USA) overnight at 4 °C. After washing with TBST (1%Tween-20 in TBS) three times, cells are incubated with Alexa Fluor 555/488-conjugated secondary antibodies corresponding to the species at room temperature for 2 h. Observation uses a fluorescence microscope (Zeiss Observer Z1, Germany) [52].

For synaptic 3D surface reconstruction, the rendered microglial surfaces mask the VGLUT1/PSD95 channels, defining synaptic puncta within the microglial surface. The “spots” function is used to detect and quantify VGLUT1 or PSD95 synaptic puncta within each microglia. The synaptic puncta per microglia is normalized to the microglial volume [53].

For mouse neuron synaptic immunofluorescence, after incubation with secondary antibodies and washing with PBS three times, slices are mounted on coverslips (Superfrost Plus, Fisher Scientific, USA) with Vectashield (H-1200, Vector Laboratories, USA). Before section permeabilization, an antigen retrieval procedure with heated citrate buffer (10mM sodium citrate (6132-04-3, Sigma Aldrich, USA), 0.05% Tween-20 (9005-64-5, Sigma Aldrich, USA), pH 8.0) is performed. Imaging is done with a confocal microscope (LSM800, Zeiss, Germany). The MetaMorph software (Molecular Devices, USA) is used to quantify synaptic puncta, their density, and average size, as well as manually quantify positive immunostaining of neuron cells [53]. To ensure objectivity and accuracy of experimental results, a double-blind design is adopted in this study. Data collection and analysis are done by a third party to avoid potential experimental bias.

Flow cytometry

Flow cytometry was utilized to assess cell death rates (Beckman Coulter, Brea, California, USA). In essence, neurons (1 × 10⁵ per well) were collected and washed in chilled PBS, then stained in the dark for 15 min using a detection reagent kit (APOAF-20TST, Sigma-Aldrich, USA). Subsequently, the precipitate was resuspended in 400 µL of binding buffer and stained with 5µL Annexin-V provided in the kit. Cell analysis was then conducted using a Novocyte flow cytometer (ACEA Biosciences Inc., San Diego, California, USA) [54].

Electrophysiological recording

Full voltage and current clamp recordings were performed using an Axopatch 700B amplifier (Molecular Devices, USA) and Digidata 1322 A (Molecular Devices, USA). To measure Action Potential (AP), a glass electrode was inserted into an intracellular solution containing 120 mM K-gluconate, 15 mM KCl, 10 mM HEPES, 4 mM Mg-ATP, 0.3 mM Tris-GTP, and 0.5 mM EGTA (pH 7.3, all from Sigma-Aldrich, USA). AP was triggered by injecting current ranging from − 150 to + 290 pA in 10 pA steps in current clamp mode. Peak numbers of AP induced by current injection were compared among neurons from different groups. For measuring spontaneous excitatory postsynaptic currents (sEPSCs), another solution filled a glass pipette with 100 mM CsCH3SO3, 20 mM KCl, 10 mM HEPES, 4 mM Mg-ATP, 0.3 mM Tris-GTP, 7 mM Tris₂₋phosphocreatine, and 3 mM Qx-314 (pH 7.3, all from respective suppliers). During the recording of spontaneous EPSCs, the membrane potential was held at −70mV. AP analysis was conducted using Igor Pro, while sEPSC analysis was performed using Mini Analysis [55].

Processing of mouse brain: Mouse brains were extracted and swiftly transferred to a chilled oxygenated solution (95% O₂/5% CO₂) containing 234 mM sucrose (Sigma Aldrich, USA), 5 mM KCl, 1.25 mM NaH2PO4 (Sigma Aldrich, USA), 5 mM MgSO4 (Sigma Aldrich, USA), 26 mM NaHCO3 (Sigma Aldrich, USA), 25 mM D-glucose (Sigma Aldrich, USA), 1 mM CaCl2 (Sigma Aldrich, USA), at pH 7.4. Subsequently, transverse cuts were made in the Hippocampal region using a tissue slicer (Leica, Germany), and the selected slices were transferred to oxygenated artificial cerebrospinal fluid (ACSF) composed of 124 mM NaCl, 2 mM KCl, 1.25 mM KH2PO4 (Sigma Aldrich, USA), 2 mM MgSO4, 26 mM NaHCO3, 10 mM D-glucose, 2 mM CaCl2, at pH 7.4. The slices were then incubated at 32 °C for 25 min, followed by at least 1 h additional incubation at room temperature (25 °C), before conducting electrophysiological recordings [56].

Electrophysiological recording parameters: As mentioned above, electrodes were inserted into the intracellular solution to trigger AP by injecting current ranging from − 20 pA to 470 pA in 10 pA increments. To standardize our testing, we recorded the resting membrane potential (RMP) and held it at − 65 mV by injecting a direct current. Input resistance (IR) was calculated by dividing the steady-state voltage response caused by a change in current (ΔV/ΔI) by the current pulse before triggering an AP, and then linear regression was applied to determine the IR value. The threshold was defined as the voltage value at which the AP slope exceeded ≥ 20 Vs − 1. AP amplitude represented the voltage difference from the threshold to the peak of the AP. AP width was the time interval between the threshold and the AP peak. The rising phase velocity was determined by the maximum value of the first derivative of the phase plot (dV/dt).

Stimulation electrode: We utilized a bipolar platinum-iridium stimulation electrode (WPI, USA) to induce AP placed in the cornu ammonis 1 (CA1) region, approximately 5 mm proximal to the stratum radiatum-CA1 border. A digital stimulator (Digitimer Ltd, UK) administered a 400 µs pulse every 15 s, ranging in amplitude from 1 mA to 3.2 mA, ensuring consistent AP generation. To maintain consistency, stimulation amplitudes were increased starting from the threshold by 1.5 times [57]. To ensure objectivity and accuracy of experimental results, a double-blind design was implemented in this study. Data collection and analysis were performed by a blinded third party to avoid potential experimental biases. The criteria for inclusion and exclusion in whole-cell recordings were as follows:

Inclusion criteria:

• 1. Seal resistance: Cells were only included for analysis if the obtained seal resistance exceeded 1 G Ω. This ensured high-quality contact between the cell membrane and the electrode, reducing background noise and current leakage.

• 2. Electrophysiological stability: The RMP of cells had to fall within the range of − 65 mV to − 80 mV and remain stable throughout the recording process to ensure physiological activity of the cells.

• 3. AP characteristics: The amplitude of the AP had to exceed 70 mV, and the phases of depolarization and repolarization needed to be distinguishable to ensure the recorded AP was biologically meaningful.

Exclusion criteria:

• 1. Seal failure: Cell recordings were excluded if the seal resistance dropped below 1 GΩ during the recording process or signs of current leakage were observed.

• 2. Abnormal electrophysiological parameters: Cells were excluded if the RMP exceeded − 80 mV or if the voltage deviated more than 10 mV from the initial value, indicating potential cellular health issues.

• 3. Unstable AP: Recordings showing significant fluctuations or changes in the morphology of AP during consecutive recordings suggested cellular physiological instability and were not considered for analysis.

RT-qPCR

Cell lysis was performed using the Trizol reagent kit (10296010, Invitrogen, Thermo Fisher, USA) to extract total cellular RNA. The quality and concentration of RNA were assessed using UV-visible spectrophotometry (ND-1000, Nanodrop, Thermo Fisher, USA). Reverse transcription was carried out using the PrimeScript™ RT-qPCR kit (RR086A, TaKaRa, Mountain View, CA, USA). Real-time quantitative RT-qPCR was conducted using SYBR Premix Ex Taq™ (DRR820A, TaKaRa) on the LightCycler 480 system (Roche Diagnostics, Pleasanton, CA, USA). GAPDH was used as the internal reference mRNA. Primers for amplification were designed and provided by Shanghai Generay Biotech Co., Ltd. The primer sequences are listed in Table S2. The relative expression of the target gene in the experimental group compared to the control group was calculated as 2^−ΔΔCt using the formula: ΔΔCT = ΔCt experimental group-ΔCt control group, where ΔCt = target gene Ct - reference gene Ct [58].

Audiogenic seizure induction test

Each mouse was acclimated for 1 min in an empty plastic cage, followed by exposure to auditory stress for 2 min with a metal bell (110–112 dB) repeated 3 times. A 2-minute rest period was given between auditory stress exposures. Audiogenic seizures were inhibited in each mouse by injecting sodium pentobarbital (sc, 400 mg/kg) (1069-66-5, Sigma Aldrich, USA) [59].

SHIRPA evaluation

We utilized a slightly modified version of the SHIRPA (SmithKline Beecham, Harwell, Imperial College, Royal London Hospital, Phenotype Assessment) protocol to assess the neurological system of mice. Essentially, mice were observed without disturbance in a transparent observation container for activity, tremors, eye closure, fur appearance, skin color, whisker appearance, lacrimation, defecation, and urination. Subsequently, the mice were transferred to another area and scored for the following parameters: transfer arousal, locomotor activity, gait, pelvic elevation, tail elevation, startle response, touch escape, and righting reflex. Then, the tail was gently restrained, and the following aspects were scored: posture passivity, trunk curl, limb grasping, and visual placing. After placing the mice on a wire mesh grid, the assessment included corneal reflex, pinna reflex, vibrissae orientation reflex, toe pinch response, and negative geotaxis. Lastly, the response to touch correction during the flip was tested, and evidence of biting and excessive vocalization was noted. A normal behavior score was assigned as 0, while an abnormal behavior was scored as 1, thereby determining the overall abnormal score for each mouse, with higher scores indicating greater abnormality. Additionally, specific functions such as motor, sensory, neurobehavior, and autonomic nervous system functions were also scored [60].

Statistical software and data analysis methods

Our study utilized R language version 4.2.1 for R programming supported by the RStudio integrated development environment, with RStudio version 2022.12.0–353. All data were processed using GraphPad Prism 8.0, and quantitative data were presented as mean ± standard deviation (Mean ± SD). Unpaired t-tests were conducted for comparisons between two groups, while one-way analysis of variance was used for multiple group comparisons. Variance homogeneity was tested using Levene’s test, and with homogeneity of variances, Dunnett’s T3 and LSD-t tests were employed for pairwise comparisons. In cases of inhomogeneous variances, Dunnett’s T3 test was used. A p-value < 0.05 was considered statistically significant for comparisons between the two groups.

Results

Identification of key immune cells—microglia, involved in post-status epilepticus seizures, through scRNA-seq

To uncover the immune mechanisms involved in the occurrence and development of seizures, we collected hippocampal tissues from 3 seizure mice (E) and 3 normal mice (N) for scRNA-seq analysis. Data integration was performed using the Seurat package. Initially, we assessed the gene numbers (nFeature_RNA), mRNA molecule counts (nCount_RNA), and mitochondrial gene percentage (percent. mt) for all cells in the scRNA-seq data. The results indicated that most cells had nFeature_RNA < 5000, nCount_RNA < 20,000, and percent.mt < 20% (Figure S2A). By filtering out low-quality cells based on the 200 < nFeature_RNA < 5000% criteria.mt < 20%, we obtained an expression matrix with 18,312 genes and 26,680 cells. Correlation analysis based on sequencing depth showed a correlation coefficient of r = − 0.22 between nCount_RNA and percent. mt, and r = 0.89 between nCount_RNA and nFeature_RNA (Figure S2B), indicating the filtered cell data were of good quality for further analysis.

Subsequent analysis of the filtered cells involved selecting highly variable genes based on gene expression variance and choosing the top 2000 variable genes for downstream analysis (Figure S2C). Cell cycle scoring was conducted using the CellCycleScoring function (Figure S2D), and data normalization was performed. Then, PCA was employed for linear dimensionality reduction based on the selected highly variable genes. The study presented a heatmap illustrating the major gene expression patterns for PC_1 - PC_6 (Figure S2E) and depicted the distribution of cells in PC_1 and PC_2 (Figure S2F). Batch effects were observed among the samples.

To mitigate batch effects and enhance the accuracy of cell clustering, the harmony package was utilized for batch correction of the sample data (Figure S3A). Additionally, the ElbowPlot was used for standard deviation-based sorting of PCs, showing that PC_1-PC_20 effectively captured the information contained in the selected highly variable genes with significant analytical relevance (Figure S3B). Post-correction results indicated successful elimination of sample batch effects (Figure S3C).

To reduce batch effects and enhance the accuracy of cell clustering, batch correction was conducted using the harmony package on the sample data (Figure S3A). Additionally, PCs were subjected to standard deviation sorting using ElbowPlot, indicating that PC_1-PC_20 adequately captured the information from the selected highly variable genes and held significant analytical value (Figure S3B). Following the correction, the results showed successful mitigation of sample batch effects (Figure S3C).

Furthermore, a non-linear dimensionality reduction of the top 20 PCs was performed using the t-SNE algorithm. Through t-SNE cluster analysis, all cells were categorized into 25 cell clusters (Figure S3D–E). Subsequently, the automatic annotation of these 25 cell clusters was carried out using the Bioconductor/R software package “SingleR,” which identified 9 cell types: Oligodendrocytes, Microglia, Endothelial cells, Astrocytes, Neurons, Fibroblasts, Macrophages, Epithelial cells, and T cells (Fig. 1A). Specifically, clusters 0, 8, 10, 11, 12, 18, 17, 21, and 22 were identified as Oligodendrocytes, clusters 1, 9, 23, 24 as Microglia, clusters 2, 4, 6, 15, 16, 17 as Endothelial cells, cluster 3 as Astrocytes, clusters 5, 14 as Neurons, cluster 7 as Fibroblasts, cluster 13 as Macrophages, cluster 19 as Epithelial cells, and cluster 20 as T cells. Furthermore, the t-SNE expression profiles of the marker genes for these 9 cell types were displayed, with Gfap as the marker gene for Astrocytes, Cx3cr1 for Microglia, Cd24a for Epithelial cells, Cdh5 for Endothelial cells, Cd3d for T cells, Rbfox3 for Neurons, Olig2 for Oligodendrocytes, Cd209f for Macrophages, and Col1a1 for Fibroblasts (Fig. 1B).

Fig. 1

Cell Annotation and Differential Analysis of scRNA-seq Data. A Visualization of cell annotation results based on t-SNE clustering, where each color represents a cell subtype; B Expression patterns of 9 cell marker genes across different cell subtypes, with darker blue indicating higher average expression levels; C Differential analysis of cell composition between Normal and seizure samples using T-test (Normal, n = 3; seizures, n = 3), cells with significant differences are highlighted by a red dashed box, “ns” indicates no difference, * indicates P < 0.05, ** indicates P < 0.01; D Percentage of different cell subtypes in each sample, represented by different colors

The distribution of these 9 cell types across 6 samples was further illustrated, and a T-test was utilized to analyze the differences in cell quantities between normal and seizure samples. The results revealed significant increases in Microglia, Macrophages, and T cells in seizure samples, while Epithelial cells significantly decreased in seizure samples (Fig. 1C, D).

Subsequently, to explore the functional differences of these significantly altered cells in the Normal and seizures groups, we utilized the R package “CellChat” to investigate pathway activities between different cells. It was observed that in the seizures group, there was an enhanced interaction between Microglia and Macrophages, as well as an increased association between T cells and Macrophages compared to the Normal group (Fig. 2A−D).

Fig. 2

Cell Communication and Pseudo-Temporal Analysis in scRNA-seq. A−B Network plots of cell communication in Normal (A) and seizures (B) samples, where the thickness of lines on the left side represents the number of pathways, and on the right side represents interaction strength; C−D Separate distinctions of interactions for each cell in Normal (C) and seizures (D) samples; E Trajectory skeleton plot colored by pseudo-temporal states, including branches (states), branch nodes (①, ②), each point representing a cell; F Trajectory skeleton plot colored by cell types

To further examine gene expression changes, single-cell pseudotime analysis was conducted using the “Monocle2” package. Within the pseudotime, cell transition trajectories displayed distinct clusters (Fig. 2E). By colorizing the pseudotime plot to determine the cellular origin tissue, a transitional relationship was identified between Microglia and Macrophages (Fig. 2F).

Through the aforementioned scRNA-seq analysis, nine cell types were identified, and notable differences in the quantities of Microglia, Macrophages, T cells, and Epithelial cells were observed between the two groups. Furthermore, cellular communication and trajectory analyses revealed a distinct connection between Microglia and Macrophages.

Identification of key genes in post-status epilepticus seizure response through integration of scRNA-seq and Bulk RNA-seq—RACK1

Microglia are macrophages within the substantial tissue of the CNS, performing two crucial functions: immune defense and maintenance of CNS homeostasis [61]. From yolk sac macrophages, microglia migrate to the brain before forming the blood-brain barrier [62].

Further analysis was conducted to identify DEGs in microglia and macrophages between seizure and normal groups, revealing 878 significantly different genes in microglia (Fig. 3A). Functional enrichment analysis indicated a significant correlation of these genes with cytoplasmic ribosomes, co-translational protein targeting to membrane, immune system processes, and cellular activation involved in immune responses (Fig. 3B). Moreover, KEGG enrichment analysis revealed significant associations of these genes with signaling pathways related to Parkinson’s disease, Alzheimer’s disease, human immunodeficiency virus type 1 infection, and human T-cell leukemia virus type 1 infection (Fig. 3C).

Fig. 3

Analysis of Differential Genes and Functional Enrichment. A Volcano plot showing differential genes in small glial cells in scRNA-seq, red dashed lines indicate genes with lower expression in the seizures group on the left side and higher expression on the right side; B−C Circos plots of functional enrichment analysis for GO (B) and KEGG (C) of differential genes in small glial cells; D Volcano plot showing differential genes in macrophages in scRNA-seq, red dashed lines indicate genes with lower expression in the seizures group on the left side and higher expression on the right side; E−F Circos plots of functional enrichment analysis for GO (E) and KEGG (F) of differential genes in macrophages; G Volcano plot of differential genes in Bulk RNA-seq, red represents significantly upregulated genes, blue represents significantly downregulated genes, and gray indicates genes with no significant difference; H−I Circos plots of functional enrichment analysis for GO (H) and KEGG (I) of differential genes in Bulk RNA-seq

In macrophages, 1293 significantly different genes were identified (Fig. 3D). Functional enrichment analysis highlighted a significant correlation of these genes with transport, establishment of localization, cytoplasmic vesicles, cellular activation, neuron morphology, and neuron death (Fig. 3E). KEGG enrichment analysis further demonstrated significant associations of these genes with phagosomes, endocytosis, human cytomegalovirus infection, Parkinson’s disease, and Alzheimer’s disease signaling pathways (Fig. 3F).

Bulk RNA-seq was performed on the Hipp of three seizure mice and three normal mice, revealing 373 significantly different genes, with 168 downregulated and 205 upregulated genes (Fig. 3G). Functional enrichment analysis showed a significant correlation of these genes with organic substance response, cholesterol biosynthetic process, cytoplasm, neuron morphology, and immunoglobulin complex (Fig. 3H). Moreover, KEGG enrichment analysis indicated significant associations of these genes with metabolic pathways, 2-oxocarboxylic acid metabolism, Parkinson’s disease, and Epstein-Barr virus infection signaling pathways (Fig. 3I).

By intersecting 7793 seizures-related genes downloaded from the GeneCards website with DEGs in microglia, macrophages, and bulk RNA-seq, eight common genes were obtained (Fig. 4A), highlighting that only the Rack1 gene exhibited consistent upregulation trends in all three expression matrices (Fig. 4B−D). To gain a deeper understanding of Rack1 expression in pseudo-time trajectories, a biased jitter plot of gene expression levels was created. This plot revealed that Rack1 exhibited the greatest expression jitter in microglia (Fig. 4E), along with a line plot indicating that Rack1 reached its peak expression in microglia (Fig. 4F). Immunohistochemical staining further confirmed the upregulated expression of RACK1 protein in the brain tissues of seizure mice (Fig. 4G). Studies have shown that RACK1 is a component of the NLRP3 inflammasome complex [63], and its signaling pathways may be involved in neural development [64]. Therefore, based on bioinformatics analysis, we identified RACK1 as a key gene in seizures.

Fig. 4

Identification of Key Genes through Integrated scRNA-seq and Bulk RNA-seq. A Venn diagrams displaying genes related to seizures downloaded from GeneCards, differential genes in microglias, differential genes in macrophages, and differential genes from Bulk RNA-seq; B−C Violin plots illustrating the expression of Rack1 in microglias (B) and macrophages (C) in scRNA-seq data (Normal, n = 3; seizures, n = 3); D Box plots showing the expression of Rack1 in Bulk RNA-seq data (Normal, n = 3; seizures, n = 3); E Biased jitter plot of Rack1 expression distribution, with cell types on the horizontal axis and gene expression levels on the vertical axis; F Presentation of gene expression changes along pseudo-temporal states, color-coded by cell types; G Immunohistochemical staining to detect the expression of RACK1 protein in mouse brain tissue, scale bar = 50 μm, the graph on the right shows the statistics of positive cells, the experiment was repeated thrice, and values are presented as mean ± standard deviation, *** indicates P < 0.001

Identification of a Novel RACK1 mutation, p. Leu206Pro, through WES

Bioinformatics analysis revealed significant upregulation of RACK1 in macrophages and microglia of the Hipp in seizure mice. Microglia play a key role as immune cells in the CNS. Subsequently, we aimed to investigate the novel genetic molecular mechanisms linking RACK1 to inflammation-induced seizures by studying genetic mutations of RACK1 in seizure patients.

WES was conducted on peripheral blood DNA from 192 seizure patients, analyzing the 8 exons of RACK1. Results identified a missense mutation (c.617T > C; p. Leu206Pro) in Exon 5 of an seizures patient (seizures NO.45), with three other patients (seizures NO.29, NO.96, NO.102) found to be heterozygous for this mutation (mutation frequency of 1.3%). To validate the findings, Sanger sequencing verification was performed (Fig. 5A).

Fig. 5

Screening for Mutations in RACK1 among Epileptic Patients. A Sanger sequencing data of RACK1 Exon 5 in 5 epileptic patients; B Predictions of the harmfulness of the c.617T > C mutation in RACK1 from PolyPhen-2 and Mutation Taster databases; C Conservation analysis of p.Leu206 in 8 species including human, rat, and mouse; D Conservation visualization of 21 amino acids using R package; E Process diagram illustrating the establishment of RACK1-p.L206P heterozygous and homozygous KI mutants using CRISPR/Cas9 editing system in HMC3; F Sanger sequencing confirmation of RACK1 mutations in allele gene cell lines

Predictions from the PolyPhen-2 and Mutation Taster databases indicated a significant harmful effect of the mutation (Fig. 5B). PhyloP and phastCons scores from Mutation Taster predicted high conservation at the site, with scores of 4.049 and 1, respectively, underscoring its conservation. Additionally, evolutionary conservation analysis of the 10 bp flanking the site across 9 species in the UCSC database indicated high conservation (Fig. 5C). Conservation analysis of the surrounding 10 amino acids using the “ggseqlogo” package in R revealed high conservation (Fig. 5D). Furthermore, this mutation was not found in the gnomAD database, signifying it as a novel finding.

Subsequently, utilizing the CRISPR/Cas9 editing system, we generated heterozygous and homozygous RACK1-p.L206P knock-in(KI) mutants in human microglia (HMC3) (Fig. 5E). To confirm the RACK1 mutations in these cell lines, Sanger sequencing of RACK1 Exon 5 was conducted on genomic DNA isolated from the cells. The T to C transversion in both strands of the codon was detected in HMC3 MUT/MUT cells, a single T to C transversion in HMC3 WT/MUT cells, with no changes observed in HMC3 WT/WT cells (Fig. 5F).

We identified a novel missense mutation, p.L206P, in RACK1 in seizure patients for the first time and successfully established homozygous HMC3 cell lines and heterozygous HMC3 cell lines harboring this mutation.

RACK1-p.L206P contributes to microglial activation and inflammatory response

One of the key pathophysiological features of seizures is neuroinflammation, involving the activation of microglia and astrocytes releasing various types of inflammatory mediators [65]. Pro-inflammatory factors can increase neuronal excitability and lower the seizure threshold in seizures [66]. Excessive activation of microglia is detrimental [67].

Initially, we observed the morphological characteristics of HMC3 cells with different genotypes under an optical microscope. Mutant cells exhibited amoeboid aberrant morphology compared to wild-type cells, with a more pronounced anomaly in homozygous mutant cells (Fig. 6A). In all subsequent cell experiments described below, microglia were activated using LPS (100 ng/mL). The expression levels of Iba-1 protein (a marker and functional indicator of microglia) were analyzed using Western blot and immunofluorescence techniques. The analysis revealed a significant increase in Iba-1 protein expression in MUT/MUT and WT/MUT cells, with MUT/MUT cells notably higher than WT/MUT cells, and the mutant cells showing amoeboid appearance (Fig. 6B−C).

Fig. 6

Evaluation of Activity and Inflammatory Activation in Cells with Different Genotypes. A Observation of morphological changes in HMC3 cells of different genotypes under an optical microscope, scale bar = 50 μm, the graph on the right shows the statistical count of irregularly shaped cells in one field of view; B Western blot analysis of Ibal-1 expression; C Immunofluorescence detection of Ibal-1 in HMC3 cells, scale bar = 25 μm, the graph on the right shows the statistics of Ibal-1 positive cells; D CCK-8 assay for cell proliferation; E MTT assay for cell viability; F ELISA measurement of pro-inflammatory cytokine levels in cell culture medium; G Griess reagent assay for nitrite content determination; H Western blot analysis of iNOS and COX2 expression levels; I−J Imaging of HMC3 cells migrating to the scratch area at different time points of 6 h, 12 h, and 24 h, scale bar = 100 μm, calculation of the percentage of migrating cells in the area using Image J software (H); K Fluorescence microscopy observation of pHrodo™ Green E. coli particles and quantitative analysis, scale bar = 25 μm. All cell experiments were repeated three times, and values are presented as mean ± standard deviation, * indicates P < 0.05, **indicates P < 0.01, *** indicates P < 0.001

Following 48 h of cultivation under the same conditions and cell density, the proliferation rate of HMC3 cells was assessed using the CCK8 assay. The results indicated the highest proliferation rate in MUT/MUT cells, followed by WT/MUT cells, and the lowest in WT/WT cells (Fig. 6D).

Additionally, cell viability was evaluated using the MTT assay, showing the highest viability in MUT/MUT cells, followed by WT/MUT cells, and the lowest in WT/WT cells (Fig. 66E). Culture media from HMC3 cells of three different genotypes were collected, and the levels of pro-inflammatory cytokines IL-1β, TNF-α, and IL-6 were measured using ELISA. The results demonstrated significantly elevated levels of pro-inflammatory cytokines in homozygous and heterozygous mutant cells compared to wild-type, with higher levels in homozygous mutant cells than in heterozygous mutant cells (Fig. 6F). Nitric oxide (NO) is a key player in inflammation and immune responses [68]. The Griess assay was utilized to estimate the release of NO by measuring nitrite levels. The nitrite content significantly increased in MUT/MUT and WT/MUT cells compared to WT/WT cells (Fig. 6G).

In this study, we utilized Western blot analysis to evaluate the expression levels of iNOS and COX2 in cells, both of which are induced by immune response and inflammation. The results revealed that cells with the MUT/MUT genotype exhibited the highest expression levels of these proteins, followed by WT/MUT cells, with WT/WT cells showing the lowest expression levels (Fig. 6H). Furthermore, activated microglia are characterized by migration and phagocytic activity. Therefore, we conducted a scratch assay to assess the migration capability of microglia with different genotypes. The number of microglia from the mutated cells migrating to the scratched area significantly increased at different time points (6 h, 12 h, and 24 h), represented as a percentage of the occupied area (Fig. 6I−J). Subsequently, we examined the phagocytic activity of microglia with different genotypes using pHrodo™ Green E. coli. The results indicated that the fluorescence intensity in MUT/MUT and WT/MUT groups was higher than in WT/WT, suggesting a significant increase in phagocytic activity in the mutant groups (Fig. 6K). Therefore, we preliminarily conclude that the RACK1-p.L206P mutation can enhance the inflammatory response of microglia, as well as their activity, migration capability, and phagocytic ability.

RACK1-p.L206P mediates neuroinflammation in microglia leading to neuronal apoptosis and abnormal discharges

The interaction between neurons and microglia is crucial in maintaining the neuroimmune system [69]. Microglia serve as key producers of cytokines and immune-related molecules in the brain, participating in neurogenesis, synaptic maturation, and neuronal circuit organization [70]. Our study revealed that the RACK1-p.L206P mutation enhances the inflammatory response and activity of microglia. We further investigated the impact of microglia with different genotypes on neurons.

To assess the toxicity of microglia on neurons, we initially transiently transfected HPPNCs with GFP (green fluorescent marker) and co-cultured them with HMC3 cells of different genotypes. Through longitudinal tracking of HPPNCs using automated microscopy, we observed their death, with the MUT/MUT group showing the highest death rate, followed by the WT/MUT group, and the WT/WT group the lowest (Fig. 7A).

Fig. 7

Exploring the Impact of Different Genotypes of HMC3 Cells on HPPNCs. A Fluorescence image of GFP-labeled HPPNCs cells, with red arrows indicating neurons that died during the experiment and white arrows indicating neurons that survived until the end of the experiment. The right image shows a statistical graph of GFP-positive cells, scale bar = 25 μm. B Quantification of Cleaved-CASP3 expression by Western blot. C Immunohistochemical detection of Cleaved-CASP3 expression, scale bar = 50 μm, with the right image showing a statistical graph of positive cells. D Detection of apoptosis in HPPNCs cells using flow cytometry, with the right image showing a statistical graph of apoptotic cell numbers. E Immunofluorescence detection of PSD95/VGLUT2 colocalization in HPPNCs cells, scale bar = 5 μm, with the right image showing a statistical graph of PSD95/VGLUT2 positive cells. F−H Whole-cell patch-clamp recordings of sEPSCs (F), analysis and comparison of sEPSC amplitudes (G) and frequencies (H). I Representative AP traces of neurons under different currents, displaying the firing frequency of AP in neurons with injected currents ranging from 0 to 150 pA. All cell experiments were repeated three times, and values are presented as mean ± standard deviation. * indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001

To further confirm whether it is the inflammation activation of microglia that leads to neuronal apoptosis, we added the culture medium of HMC3 cells with different genotypes to the HPPNCs. Subsequently, Western blot and immunohistochemical staining were employed to detect the expression of cleaved CASP3 (a typical indicator of cell apoptosis) in the cells, and flow cytometry was used to assess the apoptosis of HPPNCs. The results indicated that the mutant group significantly induced apoptosis in HPPNCs, with the homozygous group showing higher apoptosis levels than the heterozygous group (Fig. 7B−D).

To determine the changes in presynaptic and postsynaptic sites of neurons influenced by microglia, immunofluorescence staining for presynaptic VGLUT2 and postsynaptic PSD95 was performed. The results showed an increase in PSD95/VGLUT2 co-localization in the mutant group compared to the wild-type group (Fig. 7E), suggesting that the RACK1-p.L206P mutation may enhance excitatory synapses in neurons, leading to the occurrence of seizures.

We recorded sEPSC in three groups of neurons using whole-cell voltage clamp recordings, analyzing and comparing their amplitudes and frequencies. The results demonstrated that both the amplitude and frequency in the mutant group were significantly higher than in the wild-type group, with the homozygous mutant group being higher than the heterozygous mutant group (Fig. 7F−H). Additionally, we recorded the AP of neurons, analyzing their discharge frequency, and found that the MUT/MUT group exhibited the highest AP firing frequency, followed by the WT/MUT group, and the WT/WT group the lowest (Fig. 7I).

These findings suggest that the RACK1-p.L206P mutation activates the inflammatory response in microglia, leading to neuronal apoptosis and abnormal discharges.

RACK1-p.L206P mutation in mice induces epileptic responses

In the study, we validated in vitro cell experiments that the RACK1-p.L206P mutation triggers inflammation activation in microglia, leading to neuronal apoptosis and abnormal discharges. Subsequently, we further confirmed this mechanism in a mouse animal model. The corresponding mutation of RACK1-p.L206P in mice is Rack1-p.L206P. Thus, we employed CRISPR/Cas9 gene editing technology to generate Rack1-p.L206P KI mouse models, successfully establishing wild-type (L206/L206), heterozygous (L206/P206), and homozygous (P206/P206) mice with three different genotypes devoid of off-target effects. Notably, homozygous mice survived. We subsequently validated these three genotypes of mice again using Sanger sequencing (Fig. 8A).

Fig. 8

Analysis of Seizure Phenotypes in Mice with Different Genotypes. A Sanger sequencing of three different genotypes of mice. B−C Quantification of Rack1 expression in the Hipp of mice with different genotypes using RT-qPCR (B) and Western blot (C), with 6 mice in each group. D Limb clasping during tail suspension test, with the left image showing a negative response and the right image showing a positive response. E Bar graph depicting the number of negative and positive responses in tail suspension test for mice of different genotypes, with 30 mice in each group. F Steps for audiogenic seizure induction (left) and bar graph showing the number of mice with audiogenic seizures for different genotypes, with 30 mice in each group. G SHIRPA abnormality scores for 30 mice in each group. H−I Total seizure counts (H) and total duration of seizures during EEG monitoring (I) for different genotypes: P206/P206: n = 21, L206/P206: n = 15, L206/L206: n = 20. J The single spike at the onset of seizure followed by irregular spike-wave patterns. K Rapid rhythmic activities with fluctuating amplitudes. L Bar graph showing rapid rhythmic waveforms in two mutant mouse groups, P206/P206: n = 20, L206/P206: n = 13. Values are presented as mean ± standard deviation. * indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001

Initially, we extracted tissue from the Hippocampal CA1 region of the mice and assessed the expression levels of Rack1 using RT-qPCR and Western blot. The results revealed significantly higher Rack1 expression levels in L206/P206 and P206/P206 mice compared to the wild-type, with the homozygous exhibiting markedly higher levels than the heterozygous (Fig. 8B, C).

We conducted tail suspension tests to evaluate the neuropathological conditions of mice with different genotypes. Positive mice displayed limb curling when picked up by the tail, while negative mice exhibited limb extension (Fig. 8D). We recorded the number of mice showing positive reactions out of 90 mice with different genotypes and found that the highest proportion of positive reactions occurred in P206/P206 mice, followed by L206/P206, and lowest in L206/L206 mice (Fig. 8E).

We induced sound-induced seizures by exposing mice to metal ringing (110–112 dB). The results indicated that 20 P206/P206 mice exhibited epileptic responses, characterized by whole-body tonic-clonic seizures, 16 L206/P206 mice showed epileptic responses, whereas L206/L206 mice did not exhibit epileptic responses (Fig. 8F). Furthermore, we utilized the SHIRPA standard to evaluate the comprehensive behavior of the mice (normal scored as 0, abnormal as 1, specific scoring items in the experimental method), revealing a significantly higher score in homozygous mutant mice compared to heterozygous mutants and wild-type, with heterozygous mutant mice showing significantly higher scores than wild-type mice (Fig. 8G).

In our study, we implanted electrodes into the left and right frontal cortex and cerebellum of mice, utilizing an EEG system to record the brain waves of the epileptic mice observed above (P206/P206: n = 21, L206/P206: n = 15) and wild-type mice (L206/L206: n = 20) over a period of 2 weeks. We recorded the median number and total duration of spontaneous epileptic seizures. The results revealed that the median number of seizures during seizures in P206/P206 mice was 28 (interquartile range, IQR 20–30), with a median total duration of 25 min (18–33), while L206/P206 mice had a median seizure number of 15.0 (IQR 11.1–19.9) and a median total duration of 13 min (0–24). In contrast, L206/L206 mice exhibited a median seizure number of 0 (IQR 0–1) and a median total duration of 0 min (0–0.24) (Fig. 8H−I). The predominant feature of most epileptic seizures was high-amplitude rhythmic peaks occurring at a relatively constant rate or irregular intervals (Fig. 8J). A less common pattern of epileptic seizures featured high-amplitude rapid rhythmic activity, with amplitudes oscillating in a spindle-like manner (Fig. 8K). We separately tallied the number of mice in each mutant type exhibiting the frequency of occurrence of the second high-amplitude rhythmic wave, with the results showing a significantly higher frequency in homozygous mutant mice compared to heterozygous mice (Fig. 8L).

The aforementioned experiments demonstrate that RACK1-p.L206P mutant mice exhibit a seizure phenotype, with the severity of seizures higher in homozygous mutant mice than heterozygous mice.

Increased phagocytic capacity and inflammatory response of microglia in the CA1 region of mutant mice Hipp

We confirmed in vitro cell experiments the enhanced phagocytic capacity and inflammatory activation of mutant Microglia, which led us to further investigate this mechanism in vivo. Initially, we used immunofluorescence techniques to quantitatively assess the expression levels of Rack1 and Iba-1 proteins in different genotypes of mice hippocampal regions. Comparative analysis revealed a significant increase in the number of Rack1 and Iba-1 positive cells in mutant mice, particularly in the P206/P206 genotype, where the number of positive cells exceeded that of the L206/P206 genotype (Fig. 9A).

Fig. 9

Revealing the Effects of RACK1-p.L206P Mutation on Phagocytic Function and Inflammatory Response in Microglial Cells. A Immunofluorescence detection of Rack1 and Iba-1 expression in the mouse Hipp, scale bar = 100 μm, with the right image displaying a statistical graph of positive cells. B Confocal microscopy examination of phagocytic capability in microglial cells, with yellow arrows indicating phagocytic cell bodies and red asterisks representing microglial cell nuclei labeled with Hoechst, scale bar = 25 μm, and the right image showing a statistical graph of microglial cells with phagocytic ability. C−D Quantification of MerTK and Rab7 expression in the mouse Hipp using RT-qPCR (C) and Western blot (D). E ELISA measurement of TNF-α, IL-1β, and IL-6 levels in the mouse Hipp. F Immunofluorescence detection of IL-1β and Iba-1 colocalization in the mouse Hipp, scale bar = 15 μm, with the right image displaying a statistical graph of positive cells. G−H Assessment of iNOS and COX2 expression in the mouse Hipp through RT-qPCR (G) and Western blot (H). Each group consisted of 6 mice, and values are expressed as mean ± standard deviation. * indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001

Further investigation into the phagocytic function of Microglia was conducted using confocal microscopy and IMARIS image analysis tools. The experimental results demonstrated a significantly enhanced phagocytic capacity in mutant mice, with homozygous mutants showing more pronounced effects compared to heterozygous ones (Fig. 9B). Additionally, we analyzed the expression levels of the phagocytic receptors MerTK and Rab7 in the CA1 region of mice Hipp using RT-qPCR and Western blot techniques. The study results indicated that the expression levels of MerTK and Rab7 were highest in P206/P206 mice, followed by L206/P206 mice, and lowest in L206/L206 mice (Fig. 9C, D).

Subsequently, Hippocampal tissue was homogenized, and the levels of inflammatory factors TNF-α, IL-1β, and IL-6 in mice were measured using ELISA experiments. The experimental data showed that the levels of inflammatory factors were significantly elevated in P206/P206 mice compared to L206/P206 and L206/L206 mice, with L206/P206 mice displaying higher levels than L206/L206 mice (Fig. 9E). Moreover, through immunofluorescence techniques on Hippocampal slices, the co-localization of IL-1β and Iba-1 was examined, indicating a significantly higher number of IL-1β/Iba-1 positive cells in mutant mice compared to wild-type mice, with homozygous mutant mice exhibiting higher levels than heterozygous mutants (Fig. 9F). Furthermore, the expression levels of iNOS and COX2 in Hipp tissue were detected using RT-qPCR and Western blot, showing that the levels were higher in mutant mice than in wild-type mice, with significantly higher expression in homozygous mutants than in heterozygous ones (Fig. 9G, H).

In conclusion, the in vivo experimental results suggest that the RACK1-p.L206P mutation can enhance the phagocytic capacity of Microglia and induce an inflammatory response.

RACK1-p.L206P mutation leads to marked increase in neuronal excitability in mouse Hipp

To accurately assess the impact of different genotypes on neuronal excitability, we conducted comprehensive animal experimental studies. Initially, Hipp tissue slices were prepared, and membrane patch-clamp techniques were employed to record neuron AP. Remarkably, during the initial depolarization phase of stimulation, we observed that neurons in L206/L206 mice exhibited regular AP frequencies, while mutant mice showed excessive excitability, particularly where the P206/P206 genotype significantly surpassed the L206/P206 genotype (Fig. 10A). In a series of incremental depolarizing current injections ranging from 50 pA to 470 pA, we found that the number of AP emitted by mutant mice neurons was significantly higher than that of wild-type mice (Fig. 10A, B).

Fig. 10

Impact of RACK1-p.L206P Mutation on Excitability of Mouse Hippocampalocampal Neurons. A Under a maintaining voltage of − 65 mV, injection of a stepwise increasing current with an amplitude of 300 ms triggered representative AP traces; B Graph comparing AP frequencies with injected currents; C Graph depicting the maximal hyperpolarization voltage between AP induced by the injection of a depolarizing current of 270 pA; D Brief stimulation causing a synaptic AP burst; E Response of AP peak frequency to synaptic stimulation; F Area under the curve (AUC) for synaptic-induced AP. Each group consisted of six mice, with values presented as mean ± standard deviation. * P < 0.05, ** P < 0.01, and *** P < 0.001

Following a 270 pA current injection, we measured the maximum hyperpolarizing voltage between AP. In contrast, mutant mice neurons displayed a significantly depolarized maximum negative voltage afterhyperpolarization, with homozygous mice markedly higher than heterozygous mice (Fig. 10C).

Further detailed analysis of AP characteristics revealed that compared to L206/L206 mice, mutant mice displayed significantly hyperpolarized AP threshold, increased IR, and widened AP width, with homozygous mutants notably higher than heterozygous ones. However, there were no significant differences in RMP, AP amplitude, or AP rise speed (Table 1).

Table 1 Membrane properties of neurons

When Hipp tissue was briefly stimulated to induce AP (Fig. 10D), we found that the average number of AP induced in P206/P206 mice neurons was 4.2 ± 0.1, in L206/P206 mice was 3.1 ± 0.2, and in L206/L206 mice neurons exhibited lower excitability, being able to elicit an average of only 2.1 ± 0.1 AP (Fig. 10D, E). Similarly, the synaptic-induced AP duration in mutant mice neurons was longer than the burst duration in wild-type mice neurons (Fig. 10D, E). To quantitatively measure this increased burst duration, we calculated the Area Under the Curve (AUC) of AP induced by synaptic stimulation in neurons of different genotypes. Similarly, the AUC of P206/P206 mice was significantly higher than that of L206/P206 and L206/L206 mice, with L206/P206 mice also significantly higher than L206/L206 mice (Fig. 10F), further indicating greater depolarizing currents in mutant mice neurons.

Therefore, through in-depth electrophysiological experiments, we confirmed that the RACK1-p.L206P mutation leads to a significant increase in neuronal excitability in the mouse Hipp, characterized by more frequent AP bursts, increased IR and AP width, and prolonged synaptic-induced burst duration. These results provide important insights into understanding how this gene mutation affects neuronal function.

RACK1-p.L206P mutation enhances synaptic excitability

In our in vitro cell experiments, we investigated the significant impact of the RACK1-p.L206P mutation on Microglial activation and synaptic functional changes. We confirmed through in vivo experiments that mutant mice displayed notable electrophysiological alterations.

To further explore the underlying causes of AP changes, we quantitatively assessed the excitatory synaptic sites in the Hipp CA1 region using immunofluorescence techniques. Quantitative immunofluorescence analysis of excitatory synaptic markers revealed a significant increase in the expression density and average size of VGLUT1 and PSD95 proteins in all Hipp CA1 subfields (stratum oriens [SO], stratum pyramidale [SP], stratum radiatum [SR]) in mutant mice. The increase was more pronounced in homozygous mice compared to heterozygous mice (Fig. 11A-D). These data suggest that the RACK1-p.L206P mutation may primarily trigger seizure symptoms by increasing excitability at the synaptic pre- and postsynaptic levels.

Fig. 11

Exploring Alterations in Excitatory Synapses of Different Genotypes of Mice. A−D Immunofluorescent images of VGLUT1 (A) and PSD95 (C) in the hippocampalocampal CA1 region layers (SO, SP, and SR), with a scale bar of 10 μm. Quantitative statistical graphs of synaptic density and area are shown in B and D, respectively. E−F Co-localization of VGLUT1/PSD95 in the hippocampalocampal CA1 region layers (SO, SP, and SR), with white circles representing co-localization points and a scale bar of 10 μm. Quantitative statistical graphs of density and area are presented in F. Each group consisted of 6 mice, with mean ± standard deviation values. * indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001

Next, to validate the impact of Microglial activation and related inflammation on excitatory synapses, we performed immunofluorescence analysis to detect the expression of VGLUT1 and PSD95. Compared to wild-type mice, mutant mice exhibited a significant increase in the density of VGLUT1/PSD95 double-positive puncta in all layers of the Hipp CA1 region (SO, SP, and SR), with homozygous mice showing a greater increase compared to heterozygous mice (Fig. 11E, F).

Our previous studies have demonstrated that the RACK1-p.L206P mutation can enhance Microglial phagocytic activity. To further investigate whether this phagocytic activity affects the functionality of excitatory synapses, we assessed the expression of VGLUT1 and PSD95 within Iba-1-positive Microglia in the Hipp CA1 region using immunofluorescence. Through confocal imaging and three-dimensional cell surface reconstruction, we found that the expression of VGLUT1 and PSD95 within Microglia was significantly higher in mutant mice compared to wild-type mice, with homozygous mice showing a more pronounced expression.

In summary, our data support that the RACK1-p.L206P mutation not only enhances Microglial phagocytic activity but also increases synaptic excitability. These two mechanisms may cooperatively contribute to developing seizure symptoms in mutant mice. Particularly in the Hipp CA1 region of mutant mice, the increased density and size of excitatory synapses, along with the elevated expression of VGLUT1 and PSD95 within Microglia, further highlight the significant impact of the RACK1-p.L206P mutation on neuronal activity and related disorders.

Discussion

seizures are among the most common neurological events worldwide. Seizures typically occur in patients with epilepsy-a chronic disorder characterized by recurrent, unprovoked episodes. However, seizures can also occur independently, triggered by factors such as acute brain injury, infection, or status epilepticus [71, 72]. In patients experiencing seizures, dysregulation in the hippocampus may occur, involving abnormal neuronal electrical activity and irregular discharges. Microglia, a key component of the brain’s cellular network, play a crucial role in regulating neuronal homeostasis and development [73]. They are considered a specialized type of macrophage in the central nervous system (CNS) with the ability to phagocytose and clear foreign substances, as well as participate in inflammation and immune regulation [74].

In this study, we identified a novel mutation—RACK1-p.L206P—through scRNA-seq, bulk RNA-seq, and WES, providing profound insights into the molecular mechanisms underlying seizure activity, which contribute significantly to genetics, neurobiology, and neurobehavioral understanding (Fig. 12). First, scRNA-seq and bulk RNA-seq analyses identified RACK1 as a gene significantly upregulated in microglia associated with seizure activity, followed by the discovery of a novel mutation in RACK1 through WES. Furthermore, we elucidated the significant impact of the RACK1-p.L206P mutation on neuronal excitability and synaptic function, offering new perspectives on the micro-mechanisms of CNS function. Specifically, we found that the mutation enhances microglial phagocytic activity and increases synaptic excitability, both of which may contribute to the development of seizure symptoms in mutant mice.

Fig. 12

Molecular Mechanisms Underlying Seizure Occurrence Induced by RACK1-p.L206P Mutation in Mice

Moreover, the hippocampal neuron model used in this study, particularly with its focus on action potential dynamics, provides new pathways for understanding seizure mechanisms. Our research into microglial function also enhances our understanding of neuroimmune regulation, which may offer potential therapeutic strategies for neuroimmune-related disorders like seizures.

In previous studies, Xu et al. [75] demonstrated that RACK1 influences the occurrence and progression of medial temporal lobe epilepsy, while Dong et al. [76] revealed that RACK1 plays a crucial role in neuronal differentiation by regulating SCN1A expression. Building upon these findings, our study successfully identified a novel mutation, RACK1-p.L206P, closely associated with epilepsy, using wes and scRNA-seq technologies. Although many gene mutations related to seizures have been explored in past studies, this specific mutation in RACK1 is reported here for the first time. Compared to previously discovered mutations, this novel mutation offers a new perspective on the pathogenic mechanisms of seizures by inducing inflammatory activation and altering neuronal excitability in microglia [77]. This newly discovered mutation could potentially become a new target for future seizure diagnosis and treatment.

Microglia are considered the immune cells of the CNS. In this study, we found that the RACK1-p.L206P mutation significantly promoted proliferation, migration, phagocytic ability, and inflammatory activation in human microglia [78]. This finding aligns with previous studies suggesting that microglia play a key role in neurological diseases [79]. However, our research further reveals the specific role of these cells in seizures, particularly how mutant microglia promote seizure development through enhanced phagocytic activity and inflammatory activation. This provides a new direction for future exploration in seizure treatment.

Although the RACK1-p.L206P mutation appears unique in the molecular mechanism of seizures, comparing it with other known seizures-related gene mutations such as SCN1A and KCNA1 can offer a deeper understanding. For instance, SCN1A mutations typically affect sodium channel function, increasing neuron excitability [80]. In contrast, the RACK1-p.L206P mutation directly enhances the phagocytic function and inflammatory response of microglia, indirectly affecting neuron excitability. This difference may explain why the RACK1 mutation has a distinct pathogenic mechanism in seizures and potential therapeutic targets. Moreover, comparing the cellular functional changes provided by this study with known mutations strengthens the understanding of the role of RACK1 in seizure pathology and offers new directions for future research.

By utilizing CRISPR/Cas9 gene editing technology, this study successfully established cell and mouse models with the RACK1-p.L206P mutation. This technology not only improves the accuracy and efficiency of experiments but also allows for in-depth study of the mutation effects. Compared to traditional methods, gene editing technology provides greater flexibility and precision in creating seizure disease models. The successful application of this technology may drive the discovery and validation of more seizure gene mutations in the future.

This study further confirmed the close relationship between neuron excitability and seizures by observing a significant increase in neuron excitability in mutant mice. Consistent with previous research, increased neuron excitability can lead to seizures-like symptoms [81,82,83]. However, this study also provided a more comprehensive understanding of the pathological mechanism of seizures by connecting the inflammation response in microglia triggered by the mutation with changes in neuron excitability. This finding aids in precise interventions for future seizure treatments. While synaptic function changes have been extensively studied in many neurodegenerative diseases, research in seizures is relatively limited [84,85,86]. By discovering a significant increase in excitatory synaptic density and size in the Hipp CA1 region of mutant mice, this study revealed a direct link between synaptic function changes and seizures for the first time. This result may offer new opportunities for early diagnosis and treatment of seizures, especially regarding interventions targeting synaptic function.

This study identified for the first time the novel mutation RACK1-p.L206P associated with the onset of seizures and successfully constructed cell and mouse models with this mutation. We found that the RACK1-p.L206P mutation significantly promotes the proliferation, migration, phagocytic ability, and inflammation activation of human microglia, thereby inducing apoptosis in Hipp neurons and increasing the expression of AP and excitatory synaptic markers. This discovery holds significant scientific value and provides a fresh perspective on the pathological mechanisms of seizures. Understanding the role of the RACK1-p.L206P mutation in the onset of seizures in clinical practice may help to accurately assess the condition and enable early prevention and intervention of the disease. Furthermore, our research results offer new clues for developing seizure treatment strategies, such as designing drugs and therapies targeting the RACK1-p.L206P mutation.

Considering the significant impact of the RACK1-p.L206P mutation on microglial function, exploring therapeutic strategies targeting these cells is particularly crucial. For instance, developing specific inhibitors or modulators targeting microglial activation may reduce the severity of seizures by alleviating inflammatory responses. Moreover, based on the distinctive features of the RACK1-p.L206P mutation, the development of small molecule drugs or monoclonal antibodies that precisely target the altered RACK1 protein could inhibit its pathological activity. This precise molecular targeting strategy, combined with existing anti-seizure drugs, may offer a new treatment approach, especially for patients with drug-resistant seizures.

Despite the important findings of our study, there are still some limitations. First, although we induced a seizure model by injecting KA into the basolateral amygdala, our observations were focused on stages following status epilepticus. To gain a more detailed and in-depth understanding of the role of the new RACK1 mutation in seizures, we plan to extend the behavioral and EEG monitoring of mice post-KA injection in future studies. While we found that the RACK1-p.L206P mutation can induce seizures, it is unclear whether other unknown factors are involved. Additionally, our study primarily relied on mouse and cell models, and further verification is needed to confirm if these findings can be fully applicable to humans. Although we observed that the RACK1-p.L206P mutation enhances phagocytic activity in microglia and increases synaptic excitability, the detailed molecular mechanisms of this process require further investigation.

For future research, we plan to further uncover the mechanisms by which the RACK1-p.L206P mutation contributes to seizure pathogenesis and explore its interactions with other potential pathogenic factors. We also hope to gain a deeper understanding of how the RACK1-p.L206P mutation influences the phagocytic activity of microglia and synaptic excitability. On the other hand, we will explore new therapeutic strategies targeting the RACK1-p.L206P mutation, which may include drug design and gene therapy approaches. Through these efforts, we aim to advance research into the pathogenesis of seizures and provide more effective treatment options for patients with seizures.