Introduction
Epilepsy is a complex neurological disorder with a heterogeneous aetiology affecting 50 million individuals annually worldwide. Seizures can be triggered by various factors that could perturb the physiological structures or functions of the brain [
1,
2]. Unfortunately, approximately 30% of patients with epilepsy are resistant to pharmacological treatment [
3], and a significant proportion (30%) of them are classified as ‘of unknown cause’ [
4]. Even worse, no therapeutic drugs can effectively prevent the onset or progression of epilepsy. Therefore, there is an urgent need to explore new therapeutic targets and develop novel drugs that can delay or prevent the onset, inhibit the progression and mitigate the associated comorbidities of epilepsy [
5].
MicroRNAs (miRs) are small noncoding RNAs that can regulate mRNA expression levels by interacting with the 3′-untranslated regions (3′UTRs) of their target genes [
6]. Physiologically, miRs are involved in regulating various biological processes, such as neuronal development [
7], endoplasmic reticulum (ER) stress [
8] and mitochondrial function [
9]. Thus, miRs may serve as potential therapeutic targets for human diseases. Specifically, miR-211-5p regulates neuronal differentiation and viability in the brain and inhibits neurite growth and branching in vitro [
10]. Moreover, dynamic decreases in miR-211-5p expression induce hypersynchronization, leading to both nonconvulsive and convulsive seizures. Suppression of forebrain miR-211-5p further exacerbates long-lasting pentylenetetrazole-induced seizures [
11], underscoring the potential role of miR-211-5p in the development of epilepsy. Interestingly, our bioinformatics analysis revealed a decrease in miR-211-5p expression in patients with epilepsy. However, the precise mechanism by which miR-211-5p influences the occurrence and progression of epilepsy remains incompletely understood.
The degradation of haemoglobin in the brain leads to the deposition of haemosiderin, which is closely linked to neurological disorders including epilepsy [
12]. Iron overload has been considered a primary cause of refractory epilepsy resulting from haemorrhagic stroke. Chronic seizures following the injection of haemoglobin or iron salts (ferric chloride, FeCl3) into the cortex have been observed in rats. These free radicals may lead to lipid peroxidation in neuronal membranes, ultimately inducing epilepsy [
13]. These findings suggest that iron is critically involved in the pathogenesis of epilepsy. In addition, another form of programmed cell death, known as ferroptosis, is characterized by the iron-dependent accumulation of free radicals and lipid oxidation products [
14]. Unlike apoptosis and autophagy, morphological changes in ferroptosis are highlighted by an increased density of the mitochondrial membrane [
15]. Excess cytoplasmic iron can trigger lipid peroxidation, facilitating the formation of toxic lipid free radicals, as well as the initiation of ferroptosis [
16]. Of all organs, the brain is most susceptible to oxidative stress [
17], which can participate in epileptic seizure-induced neuronal cell death [
18‐
20]. In fact, reducing oxidative stress with a variety of compounds (such as antioxidants and NADPH oxidase inhibitors) might prevent seizure-induced neuronal death [
21‐
23]. Deletion of GPX4, an essential regulator of ferroptosis, has been found to induce susceptibility to epilepsy [
24]. Furthermore, the downregulation of GPX4 in epileptic mice and patient blood sample was validated in our research. Nevertheless, systematic research on the mechanism of ferroptosis in epilepsy remains lacking, while research has been performed for other nervous system diseases. Therefore, it is necessary to explore how ferroptosis and oxidative stress occur during the development and progression of epilepsy.
P2RX7 is a nonselective ligand-gated homotrimeric cation channel activated by extracellular adenosine triphosphate (ATP), which has been proposed as a possible drug target [
25]. For example, P2RX7 activation mainly occurs under pathological conditions involving high ATP release [
26,
27], such as inflammation or increased neuronal activity (e.g., during epileptic seizures) [
28]. In addition, P2RX7 receptor signaling components may serve as biomarkers for the diagnosis of temporal lobe epilepsy (TLE) [
29]. In our study, we showed that the expression of P2RX7 was significantly increased in epilepsy. Moreover, we found that the upregulation of P2RX7 in epilepsy was attributed to the downregulation of microRNA-211-5p in the preliminary bioinformatics analysis and experiments of this study. It is still unclear whether there is a regulatory relationship between P2RX7 and ferroptosis, and whether they affect the progression of epilepsy.
In this research, we identified multiple mechanisms of ferroptosis in the pathogenesis of epilepsy and clarified that miR-211-5p accounts for the increase in P2RX7. Furthermore, we confirmed that P2RX7 is a novel regulator that inhibits ferroptosis by regulating the MAPK/ERK signaling pathway during the development and progression of epilepsy. These results provide clues for new therapeutic targets and options for epilepsy patients.
Materials and methods
Patient selection and blood collection
All blood (plasma) samples from healthy controls (N = 30) and epileptic patients (N = 60) were collected at the First Affiliated Hospital of Harbin Medical University (Heilongjiang, China) with the approval of the Institutional Review Board (IRB) of the Ethics Committee. All patients had a detailed clinical assessment, and were refractory to anti-seizure medication (multidrug therapy) prior to admission. Peripheral venous blood samples were collected from each subject who was in a fasting state into a Vacutainer tube containing potassium EDTA. Then, peripheral blood mononuclear cells (PBMCs) were immediately isolated using Ficoll density gradient centrifugation. The clinical data of the epilepsy patients are presented in Additional file
1: Table S1.
Animals
Male C57BL/6 mice (aged 6–8 weeks, weighing at 20–30 g) were purchased from Harbin Medical University. All mice were housed in an indoor environment with a 12 h light/12 h dark cycle with free access to food and water. Experimental protocols for animals were approved by the Institutional Animal Care and Use Committee (IACUC) of Harbin Medical University.
Virus construction and preparation
The recombinant adeno-associated virus (AAV) vectors for transgenes P2RX7 and green fluorescent protein (GFP) were manufactured by Shanghai GeneChem Co., Ltd. (Shanghai, China). A universal scrambled sequence containing mismatched bases served as a negative control. The sequences of the control shRNA and P2RX7-shRNA were 5′-CGCTGAGTACTTCGAAATGTC-3′ and 5′-GCATCTTTGACACTGCAGACT-3′, respectively. The shRNA target sequences were inserted into GV478 lentivectors, which were transfected into 293T cells. Afterwards, viral supernatants were harvested after 48 h, with a final virus titre of 5.81E+12 v.g./ml.
Intrahippocampal injections and grouping
The mice were anaesthetized with 1% pentobarbital sodium (40 mg/kg, i.p.), fixed on a stereotactic instrument and then injected with AAV-P2RX7 (Sh-P2RX7 group), AAV-blank (Sh-Con group), PD98059 (ERK inhibitor), SB203580 (P38 inhibitor) or SP600125 (JNK inhibitor), respectively, into the hippocampus. Briefly, injections were performed using a stereotaxic frame to guide a micropipette into the hippocampus (bregma, 2.2 mm; lateral, 2.2 mm; ventral, 1.8 mm) according to the standard stereotaxic atlas. The micropipette was held in place for an additional 10 min before being slowly withdrawn. The incision was closed with sutures. Body temperature was maintained at 37 °C using a heating pad. To assess the efficiency of AAV-mediated knockdown, four mice were randomly chosen from each group and sacrificed on the 28th day after AAV injection. The hippocampus was immediately isolated and prepared for either scanning detection under confocal microscopy or Western blot analysis.
The remaining mice were stereotactically injected with kainic acid (KA, 25 mg/kg) to construct murine models of epilepsy. KA was injected slowly for 10 min and positioned in the hippocampus (AP-2.2 mm, ML-2.2 mm, V-1.8 mm). After injection, the needle was left in place for 10 min to avoid drug reflux. Two hours later, the mice received diazepam (10 mg/kg) to terminate seizures. The mice were randomly divided into eleven groups: (1) Control (n = 10, sham operation, received 1 μl PBS injection); (2) Epilepsy (n = 8, received KA injection); (3) Sh-P2RX7 + KA (n = 9, received successive injections of AAV-P2RX7 and KA); (4) Sh-Con + KA (n = 8, received successive injections of AAV-blank vectors and KA); (5) sh-P2RX7 + PD98059 + KA (n = 6, received successive injections of AAV-P2RX7, PD98059 [an ERK inhibitor] and KA); (6) sh-P2RX7 + SB203580 + KA (n = 7, received successive injections of AAV-P2RX7, SB203580 [a p38 inhibitor] and KA); (7) sh-P2RX7 + SP600125 + KA (n = 6, received successive injections of AAV-P2RX7, SP600125 [a JNK inhibitor] and KA); (8) miR-211-5p agomir + KA (n = 6, received successive injections of miR-211-5p agomir [a miR-211-5p activator, 5 nmol/g/day] and KA); (9) NC agomir + KA (n = 6, received successive injections of NC agomir [a agomir blank vectors, 5 nmol/g/day] and KA); (10) miR-211-5p antagomir + KA (n = 5, received successive injections of miR-211-5p antagomir [a antagomir inhibitor, 10 nmol/g/day] and KA) and (11) NC antagomir + KA (n = 5, received successive injections of NC antagomir [a antagomir blank vectors, 10 nmol/g/day] and KA). PD98059 (HY-12028), SB203580 (HY-10256) and SP600125 (HY-12041) were purchased from MedChemExpress (MCE). The miR-211-5p agomir, NC agomir, miR-211-5p antagomir and NC antagomir were designed by RiboBio, China. We collected the hippocampal tissue from mice 28 days after KA injection for our experiments.
Electroencephalography (EEG) measurement
After 14 days of KA injection, two bipolar tungsten electrodes (Cat No. 796000, A-M Systems) and a bipolar stainless-steel electrode were respectively implanted into the bilateral ANT and the CA1 region of the region of the right hippocampus of mouse. EEG data were collected and analyzed using NicoletOne EEG System (parameters: 1–70 Hz low- and high-frequency filter, 15 mm/s recording speed; Natus). According to the Racine scale, spontaneous recurrent seizures (SRSs) on a scale of 4–5 and their duration were observed and recorded.
Behavioural observation
Behavioural changes in epileptic mice at 90 min after KA injection were continually observed. Seizure activity was scored using the criteria described by Racine [
30]. Seizure stages were classified as follows: stage 0, no response; stage 1, ear and facial twitching; stage 2, myoclonic jerks (MJs); stage 3, clonic forelimb convulsions; stage 4, generalized clonic seizures with turning to a side position; and stage 5, generalized tonic‒clonic seizures (GTCSs) or death. Mice with at least three consecutive seizures of stage 4 or 5 were defined as fully kindled.
Transmission electron microscopy (TEM)
Mice were perfused with 2% paraformaldehyde and 2% glutaraldehyde in 0.1 mol/l sodium cacodylate buffer. The hippocampal tissue was fixed in phosphate-buffered glutaraldehyde (2.5%) and osmium tetroxide (1%). The hippocampal samples were cut, stained en bloc with 2% uranyl acetate (UA), dehydrated in acetone solutions at increasing concentrations, and embedded in an epoxy resin. The sections (70–90 nm thick) were stained with lead citrate and uranyl acetate. Ultrastructural images were then captured with a transmission electron microscope JEM1200 (JEOL, Tokyo, Japan). Images of the hippocampal mitochondria were obtained using 20× and 40× objective lenses.
Haematoxylin and eosin (H&E) and Nissl staining
Murine hippocampal tissue was fixed in 4% paraformaldehyde and embedded in paraffin. The paraffinized brains were cut into 5 μm-thick sections and the sections were dewaxed and rehydrated according to the standard protocols.
(1)
For H&E staining, the protocol was as follows: (1) mouse tissue sections were immersed into a Coplin jar containing hematoxylin and agitate for 30 s; (2) the slices were rinsed in H2O for 1 min; (3) the slices were stained with a 1% eosin solution for 10–30 s with agitation;
(2)
For Nissl staining, mouse tissue sections were immersed in staining solution for 6 min and slightly differentiated with 1% glacial acetic acid at room temperature. The rection was terminated by washing with tap water. The degree of differentiation was assessed under a microscope.
For transparent mounting, the slices were dehydrated with ethanol, washed with distilled water, clarified with xylene, sealed with neutral gum, and examined under a microscope (Olympus Corporation, Tokyo, Japan). Images of the hippocampus were captured using a 20× objective lens.
Perls’ iron staining
Paraffin slices (10 μm-thick) were cut from murine brains and incubated with xylene I and xylene II for 15 min, followed by hydration with 100%, 95%, 85%, 80%, 75% and 70% alcohol solutions for 1 min, each. Perls’ iron stain was prepared as a mixture of 80 ml 20% HCl and 80 ml 10% potassium ferrocyanide for 5 min. The tissue was incubated in the solution for 20–30 min and washed with distilled water 3 times. After eosin staining for 1 min, rapid dehydration was performed at 80%, 85%, 90% and 100% alcohol concentrations for 5 s, respectively. Finally, the slices were cleaned with xylene and sealed with resin. Images of the hippocampus were captured using a 10× objective lens.
Cell lines and culture conditions
Immortalized murine hippocampal HT22 cells [
31] (Bena Culture Collection, China) and human neuroblastoma SH-SY5Y cells, which have been widely used for studying oxidative stress [
32] and epilepsy model [
33] (ATCC, Manassas, VA, USA) were cultured in DMEM (C11995500BT, Gibco, USA) containing 10% FBS, 100 units of penicillin and 100 μg/ml streptomycin, and maintained at 37 °C in a 5% CO
2 incubator.
Cell transfection
Transfection was performed to upregulate or downregulate the mRNA expression levels of miR-211-5p and P2RX7. Briefly, miR-211-5p mimic, mimic negative control, miR-211-5p inhibitor, inhibitor negative control, siRNA-P2RX7, and siRNA-negative control were designed by RiboBio, China. According to the manufacturer’s protocol, each solution was added to Lipofectamine 3000 transfection reagent. The mixtures were equilibrated for 5 min at room temperature. Then, SH-SY5Y cells were transfected with the transfection mixture in serum-free cell medium for 48 h. The concentration of miR-211-5p mimic, mimic negative control and siRNA-P2RX7, or siRNA-negative control was 50 nM. The concentration of miR-211-5p inhibitor or inhibitor negative control was 100 nM.
Cell death assay
SH-SY5Y cells were cultured in a 6-well plate and exposed to erastin (S7242, Selleck). Erastin is a ferroptosis inducer through reactive oxygen species (ROS) and iron-dependent signaling [
34,
35]. The cells were divided into four groups: Control, Erastin (10 μM, 24 h), si-P2RX7 + Erastin, and si-Control + Erastin. After incubation with erastin for 24 h, PI and Hoechst 33342 were added at a concentration of 5 μg/ml for 10 min, each. Then, the cell death rate was measured by PI (+)/Hoechst (+). The percentage of cell death was determined as previously described [
36].
Dual-luciferase reporter assay
The plasmids containing the wild-type miR-211-5p-P2RX7 (wtLuc-P2RX7) response element and its corresponding mutant (mut-Luc-P2RX7) were purchased from Shanghai Genechem Co., Ltd. Plasmid DNA (wt-Luc-P2RX7, mut-Luc-P2RX7) and miR-211-5p mimic or miR-211-5p negative controls were co-transfected into 293T cells. Then, luciferase activity was assessed with a Double-Luciferase Reporter Assay Kit (Shanghai Genechem Co., Ltd) using the Dual-Light Chemiluminescent Reporter Gene Assay System, and normalized to firefly luciferase activity.
GEO dataset collection
Public miRNA, mRNA, and single-cell sequencing datasets were downloaded from the GEO (Gene Expression Omnibus) website (
http://www.ncbi.nlm.nih.gov/geo/) (GSE99455, GSE133554, GSE88992, GSE134697 and GSE140393). The GSE99455 [
37,
38] dataset included hippocampal miRNA profiling of 16 patients with medically intractable epilepsy and 8 postmortem healthy controls. Among 36 samples in the GSE134697 dataset [
39], 17 hippocampal and 17 neocortex samples were obtained from patients with drug-resistant temporal lobe epilepsy, whereas 2 neocortex samples were obtained from healthy subjects. Single brain cells from temporal tissues of 5 patients with temporal lobe epilepsy (TLE) were analyzed in GSE140393 [
40,
41]. Single nuclei of the CA1 area from 2 control and 2 epileptic mice were downloaded from the GSE143560 dataset [
42]. Six hippocampal samples from epileptic mice and 5 samples from control mice, each with three replicates were included in the GSE133554 dataset (
https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE133554). The GSE88992 dataset [
43] comprised 9 control hippocampus samples and 8 TLE samples from murine models.
Single-cell raw data quality control
We analyzed single-cell RNA sequencing data from the GEO database containing temporal tissues of epileptic patients (accession: GSE 143560, GSE140393) using R (version 4.1.3) or the Seurat R package (version 4.1.1). For the initial QC step, cells with expression of < 200 or > 6000 genes, mitochondrial transcript proportion > 10%, and < 200 or > 20,000 unique molecular identifiers (UMI) were filtered out of the analysis. The R package Double Finder was used to identify potential doublets [
44]. The samples were normalized using the Seurat SCTransform and combined using Seurat CCA integration workflow.
Cell type identification
The FindClusters function of the Seurat package was used to identify clusters. The FindAllMarkers function (adjusted P < 0.05 and |logFC| > 0.25) was used to determine specific genetic markers associated with each cluster. Then, T-distributed Random Neighbour Embedding (tSNE) plots were used to visualize different cells. Specifically, excitatory neurons were marked by GRIN2A, inhibitory neurons by KCNIP1, microglia by CD74, astrocytes by AQP4, oligodendrocyte precursor cells (OPCs) by VCAN, oligodendrocytes (ODCs) by MOBP, pericytes by PDGFRB, and endothelial cells by CLDN5.
Ferroptosis scores
A ferroptosis-related gene set was obtained from the FerrDb database (
http://www.zhounan.org/ferrdb/). The ferroptosis activity of an individual cell was estimated with the R package AUCell (version 1.12.0). The AUCell is a statistical method to determine if a given gene set is enriched at the top quantile of a ranking gene signature for a single cell. As such, cells expressing more genes in the gene-set have higher AUC values. The “AUCell explore Thresholds” function was utilized to select an appropriate threshold for a given gene set, and to score respective enrichment in each cell. A score greater than the threshold value was defined as a “High Ferroptosis Score”. Subsequently, cell clustering tSNE plots were generated and colour-coded according to the AUC scores, so that the distribution of ferroptosis scores might be displayed in different cellular subsets.
Differential gene expression analysis and functional enrichment
To assess differential gene expression (DEG), we executed the FindAllMarkers function integrated into the Seurat framework, employing the Wilcoxon test. Subsequently, Bonferroni correction was applied to adjust the associated p-values for multiple testing. The DEGs were then subjected to filtering, requiring a minimum log-fold change of 0.25 and a maximum adjusted p-value of 0.05. These refined DEGs were subsequently ranked based on their average log-fold change and false discovery rate (FDR).
Functional enrichment analysis of these DEGs was carried out utilizing the clusterProfiler package (v4.8.0). The gene sets used for this enrichment analysis were curated from Gene Ontology terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. Enrichment analysis was performed separately for both KEGG and Gene Ontology terms using the enrich KEGG and enrich GO functions. In Fig.
6a, we conducted enrichment analysis using differentially expressed transcription factors and differentially expressed surface proteins derived from the DEGs. This choice was made because transcription factors and surface proteins serve as essential mediators for regulating gene expression and cellular interactions with the external environment.
Western blotting (WB)
The mice were administered 1% pentobarbital sodium and decapitated. Then, bilateral hippocampal tissues were immediately collected. SH-SY5Y cells were collected after exposure. Cells and lysed tissues in RIPA buffer supplemented with protease and phosphatase inhibitors were collected. Protein concentrations were measured with BCA reagent (P0006, Beyotime Biotechnology, China). Protein samples (20 μg) were separated by SDS‒PAGE (10% separation gel) and transferred to a PVDF membrane (Merck and Co., Inc., Whitehouse Station, NJ, United States, Germany). After blocking with TBST containing 5% nonfat milk for 1 h, the membranes were incubated with Gpx4 (ab125066, 1:1000; Abcam), Hmox-1 (ab189491, 1:2000; Abcam), P2rx7 (ab259942, 1:1000, Abcam), phospho-Erk1/2 (AF1015, 1:1000; Affinity Biosciences), Erk1/2 (AF0155, 1:1000; Affinity Biosciences), phospho-p38 (AF4001, 1:1000; Affinity Biosciences), p38 (AF6456, 1:1000; Affinity Biosciences), phospho-Jnk (AF3318, 1:1000; Affinity Biosciences), and Jnk (AF6318, 1:1000; Affinity Biosciences) overnight at 4 °C. The next day, following three washes in TBST, the membranes were incubated with horseradish peroxidase (HRP)-conjugated anti-rabbit IgG (A0208, 1:2000, Beyotime) or anti-mouse IgG (A0216, 1:2000, Beyotime) at room temperature for 1 h. The protein levels were normalized to GAPDH (#5174, 1:1000, Cell Signaling Technology).
Total RNA isolation and quantitative polymerase chain reaction
A stereotactic technique was applied to inject drugs into the murine hippocampus. All male C57BL/6 mice were randomly divided into three groups: (1) Control (n = 5, sham operation, received 1 μl PBS injection); (2) Epilepsy (n = 5, received KA injection); and (3) Fer-1 (n = 5, received Fer-1 injection). Two hours after KA injection, the mice received diazepam (10 mg/kg) to terminate seizures. Then, the hippocampus was immediately isolated, and total RNA was extracted using TRIzol reagent (Invitrogen, United States) following the manufacturer’s protocol. Total RNA was quantified using a spectrophotometer (Nanodrop 2000, Thermo Fisher, USA). All RNA samples presented 260/280 nm ratios between 1.8 and 2.0.
The RNA samples were shipped to Suzhou Transcriptome Biotechnology Co., Ltd. (Suzhou, China). The mRNA was enriched by oligo(dT) beads. After concentrations were determined with an adaptor-specific q-PCR kit, equimolar samples were pooled and clustered for sequencing with HiSeq2000 (Illumina, USA).
Differential gene expression analysis was carried out using the DESeq2 package [
45]. Differences with a
p value < 0.05 and log2|fold change| > 0.585 were considered significant.
Real-time quantitative PCR
For miRNA quantification, Bulge-loop™ miRNA qRT‒PCR Primer Sets (one RT primer and a pair of qPCR primers for each set) specific for miR-211-5p and u6 were designed by RiboBio (Guangzhou, China). After drug treatment, total RNA from tissues or cell cultures was extracted using TRIzol reagent (Thermo Fisher Scientific) following the manufacturer’s protocols. Then, 1 μg of total RNA was reverse-transcribed using the Revert Aid First Strand cDNA synthesis kit (#K1622; Thermo Fisher Scientific). Real-time PCR was performed using the double-stranded DNA dye SYBR Green (#RR037A; Takara Biotechnology, Dalian, China) on an ABI7500 platform. All samples were analyzed in triplicate, and gene expression was normalized to GAPDH. The sequences of primers used to detect specific genes are listed in Additional file
1: Table S2.
Measurement of dihydroethidium (DHE)
After drug treatment, DHE (10.0 μM, 1.5 ml) was added to each well for 30 min and incubated at 37 °C. The samples were ultimately observed under fluorescence microscopy (Olympus, Tokyo, Japan).
Measurement of malondialdehyde (MDA), superoxide dismutase (SOD) and glutathione (GSH) levels
MDA, SOD and GSH levels were tested in SH-SY5Y cells and murine tissues using colorimetric assays (S0131, S0103, and S0053, Beyotime) according to the manufacturer’s protocol.
Immunofluorescence (IF) staining
Mice were anaesthetized and perfused through the left cardiac ventricle. The brains were removed and postfixed overnight in 4% paraformaldehyde. Frozen sections were prepared using a cryostat microtome with a thickness of 7-µm for IF staining. Next, tissue sections and cells were permeabilized with 0.25% Triton X-100 in PBS for 10 min, and washed 3 times with PBS before blocking with 1% BSA for 60 min. After rinsing with PBS, tissue sections and cells were incubated with primary anti-P2RX7 (28207-1-AP, 1:400, Proteintech), anti-Iba1(ab283319, 1:100, Abcam), anti-GFAP (#3670, 1:50, Cell Signaling Technology) and anti-NeuN (66836-1-lg, 1:100, Proteintech) antibodies. Each sample was incubated with a secondary antibody for 1 h at RT. After staining the nucleus with DAPI, images were captured using fluorescence microscopy (Olympus, Japan). Images of the hippocampus were captured using a 40× objective lens.
Statistical analysis
Statistical analysis was performed by GraphPad Prism 7 software (GraphPad, San Diego, CA, United States). Student’s t test was used to compare variables between two groups. One-or two-way analysis of variance (ANOVA) was used to analyze differences among groups followed by LSD post hoc comparisons when appropriate. All images are representative results from a minimum of three independent replicates with similar trends. All data are presented as the mean ± standard deviation (SD). A value of p < 0.05 (2-tailed) was considered statistically significant.
Discussion
The aetiology of epilepsy is very complex, and the pathogenesis of epilepsy remains unclear. Accumulating evidence suggests that iron is involved in epilepsy. For example, iron metabolism is related to seizures in adults with epilepsy [
56]. Additionally, dysregulation of iron accumulation and iron metabolism are associated with epilepsy and seizures after temporal lobe epilepsy [
57]. These studies underscore the importance of iron in epilepsy. However, ferroptosis, a new type of programmed cell death, is attributed to an imbalance in oxidative stress caused by iron ion accumulation. Targeting ferroptosis has been proposed as a promising therapeutic strategy for epilepsy [
58]. In addition, vitamin E [
59] and lapatinib [
60] might alleviate epilepsy by impeding ferroptosis. Nonetheless, no studies have proven how ferroptosis contributes to epilepsy and what targets can effectively regulate ferroptosis in epilepsy.
MicroRNAs are endogenous noncoding ribonucleotides. Over the past 20 years, miRs have fundamentally changed scientists’ understanding of gene regulatory networks, which is a new interdisciplinary direction in neuroscience research [
61]. The multitargeted ability of miRs provides an advantage for exploring epilepsy with complex pathophysiology. Therefore, differential expression of miRs may participate in neurological diseases [
62]. In this study, we screened epilepsy-related miRs, and identified downregulation of miR-211-5p in epileptic patients. Abnormal downregulation of miR in hippocampal tissues of epileptic mice and blood samples of epileptic patients could be verified by qRT-PCR. MiR-211-5p has been indicated in the physiology and pathology of epilepsy [
11]. Therefore, how does miR-211-5p affect the development of epilepsy? Generally, miR-211-5p could bind to several genes, including CACNA1C, P2RX7, TMTC2, JPH3 and GRM1. MiRs bind to complementary mRNAs and prevent their translation, resulting in a decrease in their protein levels. Among these genes, CACNA1C, JPH3 and GRM1 showed the same trend as miR-211-5p. In contrast, P2RX7 and TMTC2 presented a trend opposite that of miR-211-5p, so we only focused on these upregulated genes in epilepsy. While it is possible that miR-211-5p may also affect epilepsy by targeting TMTC2, our current study primarily assessed P2RX7, which displayed the most significant upregulation as measured by qRT-PCR. Therefore, we hypothesized that the upregulation of P2RX7 in epilepsy was attributed to the downregulation of miR-211-5p.
P2RX7 is a nonselective ligand-gated homotrimeric cation channel activated by extracellular adenosine triphosphate (ATP), which has been proposed as a possible drug target [
25]. In addition, P2RX7 receptor signaling components may serve as biomarkers for the diagnosis of TLE [
29]. According to data analysis and validation in the blood of patients and in vivo experiments, our study revealed that the expression of P2RX7 is upregulated. A stereotactic apparatus was applied to inject AAV into the hippocampus to knock down P2RX7 in animal models. Interestingly, P2RX7 inhibition alleviated seizures. However, the exact reasons for the upregulation of P2RX7 expression in epilepsy are not fully understood, but several hypotheses and mechanisms can explain this phenomenon, such as high ATP release [
63], increased intracellular calcium levels [
64], elevated levels of inflammation [
47] and regulation by microRNAs [
65].
Recent studies have shown a strong correlation between P2RX7 and ferroptosis in acute ischemic stroke [
66] and comorbid chronic pain and depression behaviour [
67], but according to data analysis and validation in vitro and in vivo experiments, our study was the first to find the relationship between P2RX7 and ferroptosis in epilepsy. Indeed, activation of P2RX7 leads to the activation and release of inflammatory factors from microglia [
68], as well as the modulation of neurotransmitter release from neurons [
69]. However, in our study, P2RX7 was predominantly expressed in neurons, followed by microglia, with excitatory neurons exhibiting greater ferroptosis activity than other subsets. Thus, P2RX7 alleviates seizures by inhibiting excitatory neuronal ferroptosis.
To illustrate potential signaling pathways through which P2RX7 mediates ferroptosis and alleviates epilepsy, we conducted transcriptome sequencing analysis and identified significant changes in the mitogen-activated protein kinase (MAPK) cascade. The MAPK signaling pathway plays a crucial role in various physiological and pathological processes [
70], including proliferation, migration, apoptosis, inflammation and oxidative stress [
71,
72]. The generic MAPK signaling pathway is shared by four distinct cascades, including ERK1/2, JNK1/2/3, p38-MAPK and ERK5. Highly conserved kinase cascades linking transmembrane receptors to downstream effectors are activated in response to various physiological or pathological stimuli associated with synaptic activity, plasticity and neuronal activation [
73‐
75]. For instance, CB2R could be induced by ERK and P38 to protect against the onset of epilepsy [
76]. In addition, Small et al. [
77] and Myers et al. [
78] demonstrated that JNK signaling might be involved in the pathogenesis of epilepsy. Collectively, the MAPK pathway is critically involved in epilepsy. However, whether P2RX7 regulates ferroptosis by regulating the MAPK pathway remains unknown. In our study, the MAPK pathway was activated in epileptic mice, which was reversed by inhibiting P2RX7. To investigate which cascades are more effective in regulating ferroptosis under AAV-P2RX7 conditions, specific inhibitors of ERK (PD98059), P38 (SB203580) and JNK (SP600125) were employed. As previously described, activation of P2RX7 might mediate the production of NOX2-dependent ROS by activating extracellular ERK [
79]. Consistently, blocking P2RX7 attenuated ferroptosis in the endothelium and reduced HG-induced haemorrhagic transformation after MCAO by inhibiting ERK1/2 [
80]. In our study, P2RX7 blockade regulated GPX4/HO-1 by suppressing the ERK cascade in murine models of epilepsy.
In terms of neurological diseases, downregulation of miR-211-5p was found in rat models of depression [
81] and cerebral ischemia reperfusion injury [
82]. A previous study found that the neurotherapeutic effect of empagliflozin downregulates miR-211-5p, resolving oxidative stress by suppressing the PERK/CHOP ER stress pathway in a PD model [
83]. Currently, only one study has demonstrated that dynamic decreases miR-211-5p expression induce hypersynchronization and both nonconvulsive and convulsive seizures [
11]. Our study is the first to find the relationship between miR-211-5p and ferroptosis in mice with KA-induced epilepsy. The inhibitory effects of miR-211-5p on seizures are likely associated with its inhibitory effect on ferroptosis, as (1) activation of miR-211-5p or knockdown of P2RX7 increases the GPX4 expression level and the SOD and GSH concentrations but decreases the HO-1 expression level and MDA concentration and mitochondrial membrane density, and (2) activation of miR-211-5p or knockdown of P2RX7, both of which suppress the ERK cascade in murine models of epilepsy and are closely associated with the pathogenesis of ferroptosis and oxidative stress, also alleviates seizures. These are very important findings, although other target genes of miR-211-5p or other enriched pathways need to be carefully evaluated in epilepsy models. In addition, we acknowledge the important of considering the potential impact of the vehicle (DMSO) in future studies, and we recognize that comparing Fer-1 with a solvent control can provide additional insights into the observed effects.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.