Targeting TLR4‐Depended NLRP3 Inflammasome: Sulforaphane's Mechanism in Suppressing Neuroinflammation and Seizures in Epileptic Mice

ABSTRACT

This study aimed to elucidate the role of the TLR4 pathway in the antiepileptic effects of sulforaphane (SFN). C57BL/6 mice were randomized into control, SFN, epilepsy (EP), and SFN intervention groups. SFN was administered for 7 days (25 mg/kg/day, oral gavage) to the intervention and SFN groups before epilepsy was induced. Epilepsy was induced in the EP and intervention groups using a lithium chloride–pilocarpine protocol. Behavioral and electroencephalogram (EEG) data were collected. Hippocampal tissue was examined for inflammatory markers and histopathological changes, and protein expression in the TLR4/NF‑κB/NLRP3 pathway was assessed by western blot. The involvement of TLR4 was further validated through intraperitoneal administration of the selective inhibitor TAK‑242 (3.0 mg/kg). SFN pretreatment significantly ameliorated seizure severity, mortality, pathologically elevated EEG delta wave frequency, hippocampal pro‑inflammatory cytokine levels, and neuronal damage (p < 0.05). Molecular analysis showed that SFN reduced the expression of TLR4, MyD88, p‐NF‐κB, NLRP3, and caspase‐1 (p < 0.05). SFN combined with TAK‐242 did not further enhance the reduction of these indicators, compared with SFN alone. The TLR4/NF‐κB/NLRP3 signaling pathway plays a crucial role in SFN‐mediated attenuation of neuroinflammation and contributes to its antiepileptic and neuroprotective effects.

Sulforaphane inhibits TLR4/NF‐κB/NLRP3 pathway, reduces neuroinflammation, and protects against seizures in a mouse model of epilepsy.

1. Introduction

Epilepsy (EP) is a neurological and systemic disorder characterized by recurrent spontaneous seizures attributed to abnormal or excessive neuronal activity in the brain [1]. EP affects approximately 70 million individuals worldwide. Despite long‐term pharmacological treatment, seizures and their recurrence remain uncontrolled in refractory EP, leading to physical and psychological harm. EP is therefore a global health concern; however, its underlying pathophysiological mechanisms have not yet been fully elucidated [2].

Neuroinflammation plays a central role in epileptogenesis. Inflammatory factors influence electrical activity of neurons and glial cells. In addition to inducing local inflammatory responses, epileptic seizures can trigger neuroinflammatory reactions, further exacerbating the neuronal damage [3, 4]. This creates a vicious cycle of immune inflammation, seizures, and brain injury, in which toll‐like receptor (TLR) signaling pathways are implicated [5]. Particularly, TLR4 serves as an innate immune receptor and lipopolysaccharide (LPS) sensor, modulating neuronal excitability and driving autoimmune responses that induce neuroinflammation [5]. Upon recognizing pathogen‐ or injury‐associated molecules, TLR4 activates the IKK/NF‐κB pathway via the adaptor protein MyD88. This induces the expression of NLRP3, pro‐IL‐1β, and pro‐IL‐18, providing the material basis for inflammasome assembly [6]. Activated TLR4 triggers NLRP3 oligomerization, forming a complex with p‐caspase‐1, which cleaves pro‐IL‐1β into IL‐1β. This stimulates glial cell proliferation, releases secondary inflammatory mediators such as TNF‐α and IL‐6, and induces pyroptosis, amplifying the inflammatory response. IL‐1β binds neuronal IL‐1R1 receptors and enhances the glutamatergic excitatory synaptic transmission mediated by the NMDA receptor, thereby lowering seizure thresholds [7]. It promotes endothelial expression of adhesion molecules, facilitating peripheral immune cell infiltration into the brain parenchyma, and further exacerbating neuroinflammation [8].

Sulforaphane (SFN), an isothiocyanate present in broccoli and other cruciferous vegetables as a glucosinolate precursor, exhibits anti‐inflammatory, antioxidant, and neuroprotective effects [9]. It can cross the blood‐brain barrier to reach the brain, where it has multiple functions, including inflammation and anti‐tumor effects, as well as regulation of lipid metabolism [10]. TLR4 cannot directly recognize LPS; it forms a complex with MD2, which then binds LPS. SFN targets this pathway by preferentially binding to Cys133 in MD2, thereby interfering with the TLR4/MD2 complex formation and suppressing receptor activation [11]. This direct action concurrently blocks both ligand‐induced and ligand‐independent TLR4 oligomerization [12]. Additionally, SFN indirectly suppresses TLR4 mRNA via Nrf2 activation and subsequent miR‐93‐5p upregulation [13].

Altogether, these multimodal mechanisms establish SFN as an effective TLR4 inhibitor. However, it remains unclear whether SFN reduces NLRP3 inflammasome release by regulating TLR4. To address this gap, we established an EP model with the aim to investigate the protective effects of SFN against neuroinflammation during epileptiform states and elucidate its specific mechanisms. Our findings provide novel insights into the therapeutic mechanisms of SFN in EP and neuroinflammation.

2. Experimental Materials

2.1. Laboratory Animals

SPF‐grade male C57BL/6 mice (6–8 weeks old, 18–20 g) were obtained from the Animal Center of Ningxia Medical University (SYXK(Ning)2020‐0001). Mice were housed under standard laboratory conditions with a normal light–dark cycle and provided ad libitum access to food and water.

All animal procedures were conducted in accordance with the guidelines outlined in the National Institutes of Health (NIH) guide for the care and use of laboratory animals (publication no. 8023, Revised 1978) and were approved by the Ningxia Medical University Ethics Committee (approval no.: YKLL‐2025‐0287).

2.2. Experimental Reagents and Instruments

Reagents: SFN (95% purity) was purchased from Toronto Research Chemicals (Canada); pilocarpine was purchased from Sigma‐Aldrich (USA) and lithium chloride from Maohua Tianjin University Reagent Factory.

Instruments: Microplate reader (Thermo Fisher Scientific, USA); high‐resolution scanning imaging system and cryostat microtome (Leica Microsystems, Germany).

3. Experimental Methods

3.1. Animal Grouping and Treatment

Six‐ to eight‐week‐old male C57BL/6J mice were randomly assigned to four groups: NC (blank control); SFN (25 mg/kg SFN); EP (290 mg/kg pilocarpine); SFN+EP (SFN intervention; 25 mg/kg SFN + 290 mg/kg pilocarpine) (data from SFN group are presented in the Supporting Information).

Mice in the SFN and SFN+EP groups received daily oral gavage of SFN solution, whereas NC and EP mice received a daily infusion of an equivalent volume of physiological saline. On Day 4 of treatment, electrodes were implanted into the skulls of each mouse to facilitate data acquisition during epileptic induction. On Day 7, mice in the EP and SFN+EP groups received an intraperitoneal injection of lithium chloride (130 mg/kg). After 18–20 h, an intraperitoneal injection of pilocarpine (290 mg/kg) was administered. The NC and SFN groups received an equivalent volume of physiological saline injections.

For the TLR4 pathway validation, mice were randomly assigned to four groups: EP (control); EP+SFN (SFN intervention; 25 mg/kg SFN); EP+TAK‐242 (TAK‐242 intervention; 290 mg/kg PILO + 3.0 mg/kg TAK‐242); EP+SFN+TAK‐242 (combined intervention; 290 mg/kg PILO + 25 mg/kg SFN + 3.0 mg/kg TAK‐242). TAK‐242 was administered 30 min before SFN once daily for 7 consecutive days. On Day 7, TAK‐242 was given, followed 30 min later by SFN, and another 30 min later by lithium chloride. Pilocarpine (290 mg/kg) was then injected 18–20 h later.

Behavioral alterations in mice were observed, electroencephalograms (EEG) were monitored, and behavioral data were recorded. Epileptic seizure severity was graded using the Racine scale (Table 1), with grade 4 seizures indicating successful modelling. Following sustained status epilepticus (≥ grade 4) for 2 h, mice were anesthetized via isoflurane inhalation before sacrifice..

TABLE 1.

Racine scale.

Seizure grade	Manifestation
Grade 0	No abnormal behavior.
Grade I	Facial clonus, including eye blinking, whisker twitching, and rhythmic chewing.
Grade II	Grade I manifestations plus rhythmic head nodding.
Grade III	Grade II manifestations plus forelimb clonus, without rearing.
Grade IV	Grade III manifestations plus rearing (hindlimb extension).
Grade V	Generalized tonic–clonic seizures with loss of postural control.

3.2. Indicator Determination

3.2.1. Cortical Electrode Implantation

A custom‐made electrode assembly (silver wires connected to stainless‐steel screws via a female connector) was tested for conductivity prior to use. Before surgery, the surgical sites on the scalps of the mice were disinfected. A respiratory anesthesia system was used to maintain stable anesthesia throughout the procedure, with a thermostatic device that regulated body temperature. The anaesthetized mice were secured in a stereotaxic brain mount. After adjusting the cranial level, a longitudinal incision was made along the midline and the periosteum was dissected to expose the cranial surface. A burr hole was drilled at the designated cranial site (diameter matched that of the electrode and fixation screw), with compression applied to control any bleeding. The prefabricated electrodes and fixation screws were implanted into the burr hole site. Finally, after ensuring that the silver wire was appropriately insulated and isolated, the terminal block connector was securely encapsulated using dental cement. The body temperature of each mouse was maintained post‐operatively.

Seven days after the procedure, EEG monitoring was initiated during epileptic induction using a BL420I biomedical signal acquisition and processing system (sampling rate: 50 Hz; frequency band: 250–1000 Hz). Behavioral data, latency, and time of death within 2 h were recorded. EEG data were stored for subsequent analysis. Cortical field potentials reflected the synchronized activity within the underlying and connected neuronal populations, directly quantifying seizure severity.

3.2.2. Nissl Staining

Following isoflurane anesthesia, the hearts of mice were perfused with physiological saline and 4% paraformaldehyde. The brains were subsequently harvested and fixed in paraformaldehyde at 4°C for 24 h, followed by overnight rinsing with running water. Dehydration was performed using a gradient ethanol series (70%, 80%, 95%, and 100%) for 1 h/step, after which the specimens were embedded in paraffin.

A microtome was used to prepare serial sections of 5–8 µm thickness, which were mounted onto polylysine‐coated slides and dried overnight in a 60°C oven. The sections were sequentially immersed in xylene I and II for 10 min each to remove paraffin, hydrated with graded ethanol (100%, 95%, 80%, and 70%) for 5 min each, and soaked in distilled water for 2 min. Next, the sections were stained with 0.1% tar violet for 15 min under light‐protected conditions and then rinsed with distilled water to remove excess stain.

Sections were separated using 95% ethanol for 10 s (microscopically monitored until the Nissl bodies were clearly visible), followed by dehydration in 95% and 100% ethanol for 5 min each. The sections were treated with xylene for 5 min, then mounted with neutral resin. The distribution of violet–blue Nissl bodies within neuronal cell bodies was observed under an optical microscope and images were acquired.

3.2.3. Enzyme‐Linked Immunosorbent Assay (ELISA)

After isoflurane anesthesia, mice were decapitated and the brain tissues were rapidly removed and placed on ice. The injured side of the cerebral cortex was carefully dissected using forceps and a glass dissecting needle, transferred into a clean 2 mL centrifuge tube, and processed for subsequent experiments. Next, hippocampal tissue samples were obtained from the CA1 and CA3 regions, and weighed. The samples were homogenized in phosphate‐buffered saline at a 1:9 ratio and centrifuged at 4°C for 10 min at 10 000×g. Supernatants were then collected for analysis.

Blank and sample wells were prepared in microplates. Samples and biotin‐labelled anti‐GABA antibodies were added to each well, followed by enzyme‐labelled reagents. The plates were sealed with a sealing membrane and incubated at 37°C for 30 min, during which the concentrated wash buffer was diluted. After incubation, the wells were washed five times with wash buffer and tapped dry. Color developers A and B were added, and the plates were again incubated at 37°C in the dark for another 10 min.

The hippocampal levels of key inflammatory markers TNF‐α (JONLNBIO, JL10484), IL‐6 (JONLNBIO, JL20268), and IL‐1β (JONLNBIO, JL18442) were determined using a specific ELISA kit according to the manufacturer's protocol. The concentrations of these biomarkers were calculated using ELISA Calc software, which employs standard curves for precise measurements.

3.2.4. Western Blotting

Total protein was extracted from the hippocampal tissues of mice in each group, and protein concentrations were determined using the bicinchoninic acid assay. Proteins were separated into distinct bands by 10% sodium dodecyl sulphate‐polyacrylamide gel electrophoresis, then transferred onto polyvinylidene fluoride membranes. The membranes were blocked with a rapid blocking solution for 30 min at room temperature and incubated overnight at 4°C with the following primary antibodies: mouse anti‐TLR4(1:1000, ab22048, abcam), rabbit anti‐MyD88(1:1000, ab131071, abcam), Ms anti‐NLRP3(1:2000, 68102‐1‐Ig, proteintech), rab anti‐NF‐κB‐p65(1:1000, ab76302, abcam), rab anti‐p‐NF‐κB‐p65(1:1000, TA385158, ORIGENE), Ms anti‐JNK(1:3000, 66210‐1‐Ig, proteintech), rab anti‐p‐JNK(1:1000, 88024‐1‐RR, proteintech), rab anti‐ERK(1:5000, 51068‐1‐AP, proteintech), rab anti‐p‐ERK (1:1000, 28733‐1‐AP, proteintech), Ms anti‐IkB‐Alpha (1:5000, 66418‐1‐Ig, proteintech), rab anti‐p‐IkB‐alpha (1:10000, ab133462, abcam), Ms anti‐caspase (1:5000, 81482‐1‐RR, proteintech), rab anti‐p‐caspase (1:1000, ab179515, abcam), rab anti‐beta Tubulin (1:1000, GB11017‐100, Servicebio), and Ms anti‐GAPDH (1:3000, GB15002, Servicebio).

After incubation, the membranes were washed thrice for 10 min each in Tris‐buffered saline with Tween 20 (TBST), then incubated at room temperature for 2 h with the appropriate secondary antibodies (1:500 dilution). This was followed by four TBST washes of 5 min each. Finally, the membranes were imaged using an infrared fluorescence imager, and band intensity values were analyzed using ImageJ software. The data were normalized prior to statistical analysis.

3.3. Statistical Analysis

Data are expressed as mean ± standard deviation and analyzed using GraphPad Prism 10.0 software. Comparisons among multiple groups were performed using one‐way analysis of variance, followed by Tukey's test for pairwise comparisons. Differences were considered statistically significant at p < 0.05.

4. Results

4.1. Pretreatment With SFN Conferred Significant Protective Effects Against Behavioral Seizures

The behavioral and seizure statistics of the mice following SFN administration during status epilepticus are presented in Figure 1. Mortality rates, seizure latency, survival duration, seizure frequency levels across different Racine scale grades, and total seizure frequency outcomes were compared between the EP and SFN+EP groups at 20 min post‐treatment. The SFN+EP group exhibited a significantly lower mortality rate than the EP group, along with reduced frequency of grade 5 seizures (p < 0.05). Seizure grades for each mouse are shown in Figure 1F,G.

FIGURE 1.

Effect of SFN on seizure and behavior after EP (n = 10). (A) 20‐min postictal mortality, (B) seizure latency, (C) survival time, and (D) seizure frequency at different Racine scales. (E) Total seizure frequency. (F,G) Heatmaps depicting Racine scale for EP and EP+SFN groups (Grades 3–5 indicate seizure severity and Grade 6 indicates mortality). Data are presented as mean ± SD. Error bars represent standard deviation. All replicates were biological replicates (different mouse individuals). Statistical analysis was performed using the unpaired t‐test. *P < 0.05, **P < 0.01, and ***P < 0.001 versus EP.

4.2. EEG Analysis Revealed Pronounced Differences in Cortical Activity Among the Groups

EEG data were collected and analyzed to observe physiological differences between groups. Figure 2 displays EEG amplitudes and delta wave frequency distributions. The EP group exhibited significantly increased amplitude and delta wave frequency compared with the controls. In contrast, the SFN+EP group showed statistically significant reductions in both parameters (p < 0.05). These findings indicate that SFN ameliorates seizure frequency and severity in mice with EP.

FIGURE 2.

EEG results of mice in each group (n = 3). (A) Representative EEG traces from CTRL mice. (B) Representative EEG traces from the EP group showing high‐amplitude spike‐wave discharges. (C) Representative EEG traces from the EP+SFN group demonstrating reduced epileptiform activity. (D) Quantitative comparison of average δ‐wave power among groups. Data are presented as mean ± SD. Error bars represent standard deviation. All replicates were biological replicates. Comparisons among multiple groups were performed using one‐way analysis of variance, followed by Tukey's test for pairwise comparisons. *P < 0.05, **P < 0.01, and ***P < 0.001 versus CTRL; #P < 0.05, ##P < 0.01, and ###P < 0.001 versus EP.

4.3. SFN Modulates Neuroinflammation in Epileptic Mice by Altering Pro‐Inflammatory Cytokine Levels in the Hippocampus

ELISA assays were used to measure hippocampal levels of IL‐1β, TNF‐α, and IL‐6 across groups to evaluate neuroinflammatory changes, as shown in Figure 3A–C). Compared to the control group, the EP group exhibited elevated levels of IL‐1β, IL‐6, and TNF‐α, whereas the SFN+EP group showed reduced IL‐1β and IL‐6 levels relative to the EP group. These results suggests that SFN intervention mitigates neuroinflammatory responses in epileptic mice by lowering IL‐1β, IL‐6, and TNF‐α levels.

FIGURE 3.

Analysis of hippocampal CA1 and CA3 from brain tissue of each group by ELISA and Nissl staining(n = 3). (A) IL‐1β levels, (B) TNF‐α levels, and (C) IL‐6 levels. (D) Representative Nissl staining of hippocampal CA1 and CA3 regions, scale bar = 50 µm, scale bar = 10 µm. Data are presented as mean ± SD. Error bars represent standard deviation. All replicates were biological replicates. Comparisons among multiple groups were performed using one‐way analysis of variance, followed by Tukey's test for pairwise comparisons.*P < 0.05, **P < 0.01, and ***P < 0.001 versus CTRL; #P < 0.05, ##P < 0.01, and ###P < 0.001 versus EP.

To validate this hypothesis, Nissl staining was performed on hippocampal CA1 and CA3 neurons from all groups, 2 h after seizure induction, to evaluate the neuroprotective effects of SFN, as shown in Figure 3D. In the control group, neurons in the hippocampal CA1 and CA3 regions exhibited a relatively plump appearance and were densely arranged. In contrast, the EP group exhibited shrunken neurons with indistinct outlines. However, in the SFN+EP group, these pathological changes were alleviated and neuronal death was reduced, indicating that SFN intervention mitigated EP‐induced neuronal damage and exerted a neuroprotective effect. Representative Nissl staining images of all four experimental groups are presented in Figure S1.

4.4. SFN Exerts its Anti‐Inflammatory Effects by Specifically Modulating Key Components of the TLR4/NF‐κB/NLRP3 Axis

Western blot analysis of mouse hippocampal tissue was performed to investigate the upstream signaling proteins of inflammatory factors altered in a previous study. The band patterns are shown in Figure 4A. We examined upstream pathways, including TLR4(B), ERK(C), p‐ERK(D), JNK(E), p‐JNK(F), NF‐κB (G), p‐NF‐κB (H), p‐IκBα (I), IκBα (J), p‐casp‐1 (K), casp‐1 (L), MyD88 (M), NLRP3 (N), tubulin, and GAPDH.

FIGURE 4.

SFN administration suppressed proinflammatory cytokine expression and TLR4/NF‐κB activation in the hippocampi induced by EP at 7 day post‐EP (n = 3). (A) Representative western blot bands of hippocampal TLR4, ERK, p‐ERK, JNK, p‐JNK, NF‐κB, p‐NF‐κB, p‐IκBα, IκBα, p‐caspase‐1, caspase‐1, MyD88, NLRP3, tubulin, and GAPDH in each group. Quantitative analysis of (B) TLR4, (C) ERK, (D) p‐ERK, (E) JNK, (F) p‐JNK, (G)NF‐κB, (H) p‐NF‐κB, (I) p‐IκBα, (J) IκBα, (K) p‐caspase‐1, (L) caspase‐1, (M) MyD88, and (N) NLRP3. Data are presented as mean ± SD. Error bars represent standard deviation. All replicates were biological replicates. Comparisons among multiple groups were performed using one‐way analysis of variance, followed by Tukey's test for pairwise comparisons. *P < 0.05, **P < 0.01, and ***P < 0.001 versus CTRL; #P < 0.05, ##P < 0.01, and ###P < 0.001 versus EP.

Compared with the control group, the EP group showed significantly elevated levels of TLR4, p‐ERK, JNK, NF‐κB, p‐NF‐κB, p‐IκBα, IκBα, p‐caspase‐1, caspase‐1, MyD88, and NLRP3 expression in hippocampal tissue. Meanwhile, the SFN+EP group exhibited significantly increased TLR4, JNK, NF‐κB, p‐IκBα, IκBα, and MyD88 levels compared with the control group. Notably, compared with the EP group, the SFN+EP group significantly reduced the expression of p‐NF‐κB, p‐caspase‐1, caspase‐1, MyD88, and NLRP3 levels (p < 0.05). The relative phosphorylation levels of all detected signaling proteins (ERK, JNK, NFκB, IκBα, and caspase‐1) are summarized in Figure S2A–E.

4.5. Key Pathway Components Regulated by TAK‐242

To validate the role of the TLR4 pathway in SFN‐mediated amelioration of neuroinflammation, we supplemented the interventions in each group with a TLR4 inhibitor. Western blot analysis was subsequently performed on hippocampal tissue, and the key protein bands are shown in Figure 5A. Figure 5 also displays the quantified results for TLR4 (B), p‐NF‐κB (C), p‐IκBα (D), IκBα (E), p‐caspase‐1 (F), caspase‐1 (G), MyD88 (H), and NLRP3 (I).

FIGURE 5.

SFN and TAK‐242 administration suppressed proinflammatory cytokine expression and TLR4/NF‐κB activation in the hippocampi induced by EP at 7 days post‐EP (n = 3). (A) Representative western blot bands of hippocampal TLR4, p‐NF‐κB, p‐IκBα, IκBα, p‐caspase‐1, caspase‐1, NLRP3, and GAPDH in each group. Quantitative analysis of (B) TLR4, (C) p‐NF‐κB, (D) p‐IκBα, (E) IκBα, (F) p‐caspase‐1, (G) caspase‐1, (H) MyD88, and (I) NLRP3. Data are presented as mean ± SD. Error bars represent standard deviation. All replicates were biological replicates. Comparisons among multiple groups were performed using one‐way analysis of variance, followed by Tukey's test for pairwise comparisons. *P < 0.05, **P < 0.01, and ***P < 0.001 versus EP.

Compared with the EP group, the EP+SFN group exhibited reduced levels of TLR4, p‐NF‐κB, p‐IκBα, p‐caspase‐1, MyD88, and NLRP3. The EP+TAK‐242 group showed decreased p‐IκBα, p‐caspase‐1, and NLRP3. while the EP+SFN+TAK‐242 group exhibited reductions in TLR4, p‐NF‐κB, p‐IκBα, p‐caspase‐1, MyD88, and NLRP3 (p < 0.05). The comparison of protein phosphorylation levels among the four groups is presented in Figure S2F,G.

4.6. SFN Alleviates Neuroinflammation Via the TLR4 Pathway

Following TLR4 inhibition with TAK‐242, we used ELISA assays to measure hippocampal IL‐1β, TNF‐α, and IL‐6 levels across groups, and the results are presented in Figure 6A–C.

FIGURE 6.

TAK‐242‐mediated therapeutic inhibition of the TLR4/NLRP3 Axis alters neuroinflammatory outcomes in EP mice revealed by ELISA and Nissl staining (n = 3). (A) IL‐1β levels; (B) TNF‐α levels; (C) IL‐6 levels. (D) Representative Nissl staining of hippocampal CA1 and CA3 regions, scale bar = 50 µm, scale bar = 10 µm. Data are presented as mean ± SD. Error bars represent standard deviation. All replicates were biological replicates. Comparisons among multiple groups were performed using one‐way analysis of variance, followed by Tukey's test for pairwise comparisons. *P < 0.05, **P < 0.01, and ***P < 0.001 versus EP+ vehicle; #P < 0.05, ##P < 0.01, and ###P < 0.001 versus EP + TAK‐242 + vehicle.

Compared to the EP+vehicle group, all intervention groups showed decreased cytokine levels. However, compared to the EP+TAK‐242 group, the EP+SFN+vehicle group differed in IL‐6 and TNF‐α levels, while the EP+SFN+TAK‐242 group exhibited reduced IL‐1β and IL‐6 levels. Approximately 2 h after epileptic induction, Nissl staining of CA1 and CA3 neurons was used to evaluate the neuroprotective effects of SFN, as shown in Figure 6D. The EP group exhibited shrunken, indistinct contours, whereas the other groups showed relatively plump, tightly arranged neurons. Treatment with SFN and TAK‐242 significantly reduced neuronal death in EP mice, indicating that the effects of SFN align with those of TAK‐242 (p < 0.05).

4.7. EEG Evidence Following TAK‐242 Co‐Treatment Demonstrating Critical TLR4 Involvement in the Antiepileptic Effects of SFN

Figure 7 presents the EEG amplitude and frequency measurements obtained from each TAK‐242‐treated group using the same methodology. Compared with the EP+vehicle group, all other groups showed significantly reduced delta wave frequency. However, no significant differences were observed between the EP+TAK‐242+vehicle, EP+SFN+vehicle, and EP+SFN+TAK‐242+vehicle groups. The TLR4 inhibitor effectively suppressed abnormal EEG activity in epileptic mice, significantly reducing the delta wave frequency (a hallmark of epileptic activity). This result further confirms that SFN improves epileptic electrophysiological disturbances by targeting the TLR4 pathway. Compared with previous EEG findings, SFN and TLR4 inhibitors demonstrated equivalent effects, indicating the involvement of the TLR4 pathway in SFN‐mediated modulation of abnormal epileptic EEG (p < 0.05).

FIGURE 7.

EEG results of mice in each group (n = 3). (A) Representative EEG traces from the EP + vehicle group. (B) Representative EEG traces from the EP + SFN + vehicle group. (C) Representative EEG traces from the EP + TAK‐242 + vehicle group. (D) Representative EEG traces from the EP + SFN + TAK‐242 + vehicle group. (E) Quantitative comparison of average δ‐wave power across experimental groups. Data are presented as mean ± SD. Error bars represent standard deviation. All replicates were biological replicates. Comparisons among multiple groups were performed using one‐way analysis of variance, followed by Tukey's test for pairwise comparisons. *P < 0.05, **P < 0.01, and ***P < 0.001 versus EP+ vehicle.

4.8. Regulation of the TLR4 Pathway With TAK‐242 Produces Significant Anti‐Seizure Effects

The behavioral and seizure outcomes observed in mice treated with the TAK‐242 inhibitor in the EP and SFN+EP groups are presented in Figure 8. These data include seizure latency, survival duration, seizure frequency levels across different Racine scale grades, and total seizure frequency for the EP+vehicle, EP+SFN+vehicle, EP+TAK‐242, and EP+SFN+TAK‐242 groups. Compared with the EP+vehicle group, all other groups exhibited increased seizure latency and survival duration, along with reduced seizure frequency (total as well as grades 4 and 5). Seizure grades for each mouse across the four intervention groups are presented in Figure 8E–H.

FIGURE 8.

Effect of TAK‐242 on seizure and behavior after EP and EP+SFN (n = 5). (A) Seizure latency, (B) survival time, (C) total seizure frequency, and (D) seizure frequency at different Racine scales. (E–H) Heatmaps depicting Racine scale for groups of EP+vehicle, EP+SFN+vehicle, EP+TAK‐242, and EP+SFN+TAK‐242 (Grades 3–5 indicate seizure severity and Grade 6 indicates mortality). Data are presented as mean ± SD. Error bars represent standard deviation. All replicates were biological replicates. Comparisons among multiple groups were performed using one‐way analysis of variance, followed by Tukey's test for pairwise comparisons. *P < 0.05, **P < 0.01, and ***P < 0.001 versus EP+ vehicle.

5. Discussion

Building upon the known anti‐neuroinflammatory effects of SFN, this study extended its investigation to a lithium‐pilocarpine‐induced epileptic mouse model. The results revealed that SFN treatment alleviated EP‐induced abnormalities, including EEG disruptions and neuronal damage. These comprehensive neuroprotective effects are consistent with reduced expression of the TLR4/MyD88/NF‐κB/NLRP3 signaling pathway [14].

To determine whether the TLR4/NF‐κB pathway is essential for mediating the anti‐inflammatory and neuropathological effects of SFN, we assessed the effects of the TLR4 inhibitor TAK‐242 in EP and SFN‐treated mice. Both TAK‐242‐ and SFN‐treated EP groups exhibited marked protective effects compared with the untreated EP group, while no additive effect was observed with their combined administration. We consider this as strong evidence that SFN exerts its beneficial effects via the TLR4 pathway [15].

Our behavioral and EEG findings are consistent with those of previous studies, indicating that SFN possess anticonvulsant properties [16]. Here, we examined changes in neuroinflammasome activity following SFN intervention in EP and validated these findings using known pathway inhibitors. The post‐seizure release of damage‐associated molecular patterns (DAMPs), such as ATP and HMGB1 [1], activates TLR4 as an endogenous ligand [17, 18]. In line with this mechanism, we confirmed the significantly increased expression of TLR4 and its downstream adaptor protein MyD88 in the EP group, leading to amplification of inflammatory signaling cascades [19]. Activated NF‐κB subsequently initiates transcription of genes that encode NLRP3 and pro‐IL‐1β, laying the groundwork for NLRP3 inflammasome assembly [20]. Our data demonstrate that SFN treatment effectively reduced the expression at multiple key nodes within this pathway, including TLR4, MyD88, p‐NF‐κB, and NLRP3 [14].

Regarding the role of caspase‐1, our study showed that SFN reduced the expression of caspase‐1. Activated caspase‐1 not only processes pro‐IL‐1β into mature IL‐1β [21], which has potent pro‐epileptogenic effects, but may also cleave GSDMD to induce pyroptosis, directly causing neuronal death and amplification of inflammation [22]. The reduced expression of caspase‐1 by SFN therefore likely represents a central mechanism underlying the concurrent alleviation of inflammation and neuronal injury. The ELISA and Nissl staining results in this study corroborate this interpretation, demonstrating reduced hippocampal IL‐1β levels alongside preservation of neuronal morphology following SFN treatment. Supplementary analysis of protein phosphorylation further revealed elevated p‐ERK/ERK and p‐NF‐κB/NF‐κB levels during the epileptiform state, which were attenuated by SFN intervention.

Recent studies have indicated that numerous natural compounds possess the capacity to modulate TLR4 signaling, thereby exerting anti‐inflammatory and neuroprotective effects. For instance, curcumin, the primary active component of turmeric, directly binds to the TLR4/MD2 complex to inhibit LPS‐induced NF‐κB activation and release of downstream inflammatory mediators [23]. This process regulates microglial activation, thereby mitigating LPS‐induced neuroinflammation.

Resveratrol, commonly found in grapes and blueberries, inhibits NLRP3 inflammasome activity, reducing downstream inflammatory mediators and inducing the production of antioxidant enzymes. This has been shown to improve cognitive impairment in Alzheimer's disease and prevent neuroinflammation [24]. Another flavonoid compound, quercetin, widely present in apples and onions, exerts anti‐inflammatory effects by inhibiting the NF‐κB and p38 MAPK pathways [25], and has also been shown to have anti‐neuroinflammatory effects in Alzheimer's disease [26].

Furthermore, epigallocatechin gallate (EGCG) in green tea has been reported to reduce TLR4, p‐IKK/IKK, and p‐NF‐κB/NF‐κB expression levels, thereby inhibiting NLRP3 activation in rat primary microglia and mouse hippocampal cells, thus diminishing microglial inflammation [27]. Together, these studies support the hypothesis that TLR4 represents a promising therapeutic target for natural compound‐based interventions.

To establish the role of TLR4 in the action of SFN, we employed a pharmacological inhibition strategy. TAK‐242, a specific inhibitor of TLR4 signaling, binds to the intracellular domain of TLR4, thereby blocking its interactions with downstream molecules. Although our primary focus is the direct regulation of TLR4 by SFN, the potential counter‐regulatory role of Nrf2 cannot be overlooked. Compared with previous studies describing indirect regulation of TLR4 via Nrf2 counter‐regulation, our findings demonstrate that TAK‐242 alone confers significant protective effects at behavioral, electrophysiological, inflammatory, and molecular levels. This indicates that, even in the presence of other potential targets (e.g., Nrf2), reduced expression of the TLR4 pathway alone is sufficient to produce the primary antiepileptic effects [10, 13, 28].

Supporting Information detailing changes in protein phosphorylation levels following TAK‐242 treatment demonstrate that both SFN and TAK‐242 alleviated inflammatory protein phosphorylation in epileptic mice. Furthermore, the absence of additive benefits from combining SFN and TAK‐242 strongly supports their operation within the same linear pathway, confirming that SFN is an upstream inhibitor of the TLR4 pathway. Previous studies provide molecular insights supporting this conclusion. Through molecular docking and mutagenesis experiments, Koo et al. discovered that SFN preferentially forms a covalent bond with the Cys133 residue of the TLR4 co‐receptor MD2. This disrupts TLR4/MD2 complex formation, thereby blocking the signal transduction triggered by LPS or endogenous DAMPs [11]. While our study further validated the functional relevance of TLR4 inhibition using TAK‐242, limitations remain in the molecular component.

Collectively, these findings indicate that SFN attenuates neuroinflammation and exerts antiepileptic effects by suppressing the TLR4/NF‑κB/NLRP3 signaling pathway, as summarized in Figure 9.

FIGURE 9.

Schematic diagram of the neuroinflammation protective effect of SFN in EP.

This study also has several additional limitations. First, we acknowledge that this study did not provide direct evidence of SFN binding to MD2/TLR4 (e.g., molecular interaction experiments), primarily due to the procurement cycle of relevant antibodies and budget constraints at this stage. Although we functionally confirmed the regulatory effect of SFN on the TLR4 pathway by detecting downstream molecule expression and pharmacological validation with TAK‐242, we were unable to directly examine the effect of SFN on TLR4‐MD2 complex binding. At the same time, the possible involvement of the Nrf2 pathway in the anti‐inflammatory effects of SFN cannot be completely excluded, and we plan to further investigate this in future studies.

Second, our primary focus was on inflammatory changes within the hippocampal region, whereas other brain areas within the epileptic network are equally significant. Given our investigation of temporal lobe EP, inflammatory alterations in cortical regions may also warrant consideration. However, this study lacked cortical EEG recordings and Nissl staining data from the cortex and other brain areas, thus failing to comprehensively reflect behavioral and morphological changes in mice following EP.

Third, the systemic administration method employed cannot entirely exclude the potential effects of SFN on the peripheral immune system, which may introduce ambiguous influences on the experimental accuracy [29]. Future studies should utilize hippocampal TLR4 knockout mice or adopt local administration approaches to confirm the direct target effects of SFN within the central nervous system [30].

Fourth, in designing the experiment, we referenced the existing SFN safety data and selected a daily oral dose of 25 mg/kg as the study benchmark. Unlike other studies, we did not employ a graded‐dose intervention approach, and thus failed to investigate the effects of SFN on the TLR4 pathway at different concentrations. Future work will address this limitation by systematically comparing multiple SFN doses [31, 32, 33].

Fifth, our intervention protocol primarily constitutes a proof‐of‐concept design aimed at rapidly and definitively assessing the acute antiepileptic and neuroprotective potential of targeting the TLR4 pathway. In contrast to the conventional clinical scenario of prophylactic SFN administration as a dietary supplement, this study provided fundamental evidence of the immediate effects of TLR4 inhibition, validating its role in the epileptic seizure process. Concurrently, we implemented a one‐week SFN intervention prior to EP modelling to exclude acute dosing effects on the physiological state of the mice. Future research should employ commercially available SFN supplements in chronic models that more closely resemble clinical practice, and explore the impact of long‐term prophylactic administration (such as sustained SFN intake) on seizure occurrence and related indicators.

6. Conclusion

This study systematically elucidates the mechanism by which SFN exerts anticonvulsant, anti‐inflammatory, and neuroprotective effects in an EP model, primarily through the regulation of the TLR4/NF‐κB/NLRP3 signalling pathway. This finding is supported by complementary findings using the TLR4 inhibitor TAK‐242. SFN regulates the TLR4 pathway, thereby reducing the production and release of the key pro‐inflammatory cytokine IL‐1β and reducing the expression caspase‐1‐mediated inflammatory responses. These findings not only deepen our understanding of the neuroinflammatory mechanisms underlying EP, but also provide compelling preclinical evidence for developing SFN or other drugs targeting the TLR4 pathway as adjunctive therapeutic strategies for EP management.

Funding

We gratefully acknowledge the Natural Science Foundation of Ningxia (2023AAC03549) for funding.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: mnfr70510‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.