Background
In recent years, a growing body of evidence has shown that inflammatory response plays a role in epileptogenesis [
1‐
4]. The Toll-like receptor 4 (TLR4) signaling pathway is the most studied pathway. TLR4, as a pattern recognition receptor, can promote innate immune response and mediate inflammatory responses to lipopolysaccharide [
5]. In central nervous system (CNS), TLR4 is expressed in astrocytes, oligodendrocytes, microglia, and neurons. TLR4 expression is low in the resting state. In some pathological conditions, TLR4 expression in the cytoplasm increases and TLR4 proteins are transferred to the cell membrane [
5]. High mobility group box-1 (HMGB1) is one of the important endogenous activators of TLR4. After TLR4 activation, the downstream pathway is activated primarily through a MyD88-dependent pathway. An activated TLR4 recruits MyD88, thereby causing ubiquitination of TRAF6 and NF-κB inflammatory pathway activation. An activated NF-κB moves into the nucleus and promotes the transcription of pro-inflammatory cytokine IL-1β and TNF-α [
5]. Previous studies showed that the TLR4 pathway plays a role in ischemic brain injury, brain trauma, CNS infection, neurodegenerative disease, and brain autoimmune inflammation [
5‐
7], whereas only a preliminary understanding of the role of TLR4 in epileptogenesis has been obtained in recent years [
8‐
10]. Focal cortical dysplasia (FCD) is one of the major causes of refractory epilepsy and is the most common histopathological type in children with lesional epilepsy [
11,
12]. In this study, we collected FCD type II and peri-FCD paired samples of surgically operated tissues from eight children with refractory epilepsy. We examined TLR4 and its endogenous agonist HMGB1 and detected the expression of TLR4/MyD88 complex, ubiquitination of TRAF6, phosphorylation of IKK, phosphorylation of IκB-α, phosphorylation of NF-κB p65, and pro-inflammatory cytokines IL-1β and TNF-α in the downstream pathways. We attempted to determine whether the HMGB1-TLR4 pathway is upregulated in FCD type II lesions and to provide further evidence on the role of inflammatory response in epileptogenesis.
Methods
Selection of patients
The patients met the following inclusion criteria: (1) diagnosed as having refractory epilepsy according to the criteria defined by the International League Against Epilepsy; (2) focal cortical dysplasia was suggested in brain magnetic resonance imaging (MRI): focal changes in the cortical thickness, signal increase (mainly on T2FLAIR and T2WI sequences), gray-white matter blurring, and the presence of a “transmantle” sign; (3) other etiologies potentially causing epilepsy were excluded; (4) lesions were confirmed to be FCD type II (a or b) in postoperative pathology; (5) peripheral zone of resected lesions were microscopically confirmed without dysplastic neurons and balloon cells; and (6) the size of lesions was large enough to satisfy the need of both clinical pathological examination and the research. Forty patients diagnosed with FCD type II and underwent lesion resection were collected from January 2016 to April 2017 in the Children Epilepsy Center, Peking University First Hospital. Of these patients, 17 met the inclusion criteria. We randomly selected eight cases, including three cases of FCD type IIa and five FCD type IIb.
The following clinical data were collected: (1) demographic data; (2) information associated with epilepsy, particularly age of onset, seizure type, seizure frequency, status epilepticus, and antiepileptic drugs; and (3) information associated with lesion resection, particularly age of operation, location of FCD, and pathological type of FCD. Detailed clinical data are shown in Table
1.
Table 1
The clinical data of eight children with FCD type II
Gender | Female | Male | Male | Male | Female | Female | Male | Female |
Age of onset | 6 months | 1 year | 40 days | 3 years | 7 months | 4 months | 6 months | 1 year and 10 months |
Seizure type | FS | FS | FS | FS | FS | Spasm | FS, atypical absence | FS |
Seizure frequency | 10/day | 1–2/day | 2–10/day | 1/week | 1–4/day | 3–4/day | 2–4/day | 3–10+/day |
Status epilepticus | None | None | None | None | None | None | EPC | None |
Antiepileptic drugs | VPA, LEV, TPM, VGB | OXC, LTG, PB, VPA, TPM, CZP, LEV | VPA, OXC, TPM, PB, LEV, VGB | LEV, CBZ, LTG, VPA, TPM | VPA, OXC, PB, LTG, LEV, TPM | VPA, TPM, OXC | VPA, LEV, CZP, OXC | VPA, OXC, TPM, LTG, NZP |
Age of operation | 1 year and 2 months | 7 years and 3 months | 2 years | 4 years and 2 months | 3 years and 6 months | 1 year and 2 months | 7 years and 3 months | 13 years |
Location of FCD | Left temporal | Right parietal | Left parietal | Left frontal | Left central | Right frontal | Right frontal | Right temporal |
Pathological type of FCD | FCD IIa | FCD IIa | FCD IIa | FCD IIb | FCD IIb | FCD IIb | FCD IIb | FCD IIb |
Acquisition of brain tissue samples
During the operation, brain tissue sample inside the FCD and in the peri-lesional zone were obtained respectively; both were 1 cm3 in size. Each sample was divided into two parts. One part was stored in liquid nitrogen and prepared for protein and total RNA, whereas the other part was fixed in 4% buffered formalin and prepared for the immunofluorescence test.
The study was approved by the clinical research ethics committee of Peking University First Hospital. Written informed consents were obtained from the parents.
Detection of HMGB1-TLR4 pathway in FCD and peri-FCD
Detection of HMGB1-TLR4 pathway in a tissue sample
Detection of HMGB1-TLR4 pathway in glia and neurons
Immunofluorescence double staining was carried out to detect the expression and distribution of HMGB1 and TLR4 in neurons, astrocytes, and oligodendrocytes. The brain tissue was fixed in 10% buffered formalin for 24 h. In all cases, including lesion and peripheral zone, a representative formalin-fixed, paraffin-embedded tissue block was performed. Paraffin-embedded tissue was sectioned at 4 μm and mounted on precoated glass slides. The tissue sections were blocked with 5% normal goat serum for 1 h at 37 °C. After that, the sections were incubated with the following primary antibodies: HMGB1 (rabbit monoclone, 1:200; mouse monoclone 1:50 Abcam), TLR4 (mouse monoclone, 1:100, Abcam), GFAP (mouse monoclone, 1:200, CST; rabbit monoclone, 1:200, CST), NeuN (rabbit monoclone, 1:500, Abcam; mouse monoclone, 1:500, Abcam), and Olig2 (rabbit polyclone, 1:100, CST) at 4 °C for overnight. On the following day, after rinsing with phosphate-buffered saline (PBS), sections were incubated with Alexa Fluor 568 and Alexa Fluor 488 (anti-rabbit IgG or anti-mouse IgG; 1:500) for 2 h at room temperature. The cell nuclei were labeled using Hoechst 33342.
Quantitative analysis was performed for the percentage of double-labeled positive cells. Briefly, five representative non-overlapping fields of the tissues were captured and digitized with a laser scanning confocal microscope (Olympus). We counted the total number of neurons (NeuN-positive), astrocytes (GFAP-positive), and oligodendrocytes (Olig2-positive) in each field. HMGB1, TLR4-positive in neurons, astrocytes, and oligodendrocytes were subsequently counted, and the percentage of TLR4-positive cells marked specific marker and the percentage of HMGB1-specific markers/Hoechst colocalization were calculated.
Correlation between TLR4 expression and clinical variables
After comparing FCD and peri-FCD of each patient, we investigated the correlation between the TLR4 expression levels in the lesions and clinical variables, including disease duration of epilepsy and seizure frequency. The disease duration of epilepsy was calculated by subtracting age of onset from age of operation. Seizure frequency was defined as the average number of seizures per day during the disease course. All experiments were performed independently for three times.
Statistical analysis
Statistical analysis was performed with SPSS 20.0. All data were expressed as mean ± SEM of the mean. Western blot, real-time PCR, and co-immunoprecipitation analyses between FCD and peri-FCD in the same case were performed using paired-sample t test. The correlation between TLR4 expression level and clinical variables was assessed using the Spearman’s rank correlation test. The immunofluorescent staining results were analyzed using the nonparametric Wilcoxon rank-sum test. A P value of 0.05 was considered statistically significant.
Discussion
Few previous studies suggested an upregulation of HMGB1-TLR4 pathways in epileptic brain tissues [
8‐
10]. As an important pattern recognition receptor in the innate immune response, TLR4 is mainly expressed in the astrocytes, oligodendrocytes, microglia, and neurons in the brain [
5]. In the resting state, TLR4 has low expression level in the brain, although it can be elevated in some pathological conditions (infection and tissue damage). TLR4 can be activated by exogenous pathogen-associated molecular patterns, such as lipopolysaccharide and endogenous damage-related molecular patterns, including HMGB1. And activated TLR4 can subsequently activate the downstream NF-κB pathway and promote the synthesis and release of IL-1β and TNFα. These processes lead to inflammatory response. In the resected lesion brain tissue of adult hippocampal sclerosis, Maroso et al. [
8] showed that the expression level of TLR4 in the astrocytes and neurons was higher than that in normal tissues, and the expression of HMGB1 in astrocytes was increased in the cytoplasm, suggesting that HMGB1 in astrocytes was released from the nucleus to the cytoplasm. Zurolo et al. [
9] found that in the lesion tissue of focal cortical dysplasia, cortical tubers of tuberous sclerosis, and ganglion glioma, TLR4 expression in the neurons and astrocytes is higher than that in normal human autopsy tissues and HMGB1 is significantly expressed in astrocyte cytoplasm. In this study, we used the surgically resected brain tissues form epileptic children with FCD II and compared them with the peripheral tissues from the same children. We found that TLR4 endogenous agonist HMGB1 was released from the nucleus into the cytoplasm. TLR4, TLR4/MyD88 complex, and ubiquitinated TRAF6 expression were upregulated, and the downstream NF-κB inflammatory pathway was activated, thereby increasing the expression of pro-inflammatory cytokines IL-1β and TNFα. Our findings provided a strong evidence for the upregulation of the HMGB1-TLR4 pathway in the epilepsy lesion, and this upregulation led to the release of downstream pro-inflammatory factors.
How was TLR4 pathway upregulated in epilepsy? Although TLR4 is involved in pathogen recognition and activated during infections, the HMGB1-TLR4 axis might underlie chronic epilepsy in scenarios without frank immune activation or infection. HMGB1 is a nonhistone chromosome-binding protein that is abundantly present in the nuclei of eukaryotic cells. When cells are damaged or stressed, HMGB1 is released from the nuclei of the injured or necrotic cells to the cytoplasm and extracellular membrane [
14]. The release of HMGB1 can be considered a dangerous signal from the damaged or stressed cells to alert tissue of the status of acute or persistent damage. In this study, we confirmed that HMGB1 is released from the nucleus to the cytoplasm and extracellular membrane in neurons and astrocytes inside FCD lesions. HMGB1 was released to the extracellular membrane, and it interacted with TLR4 and activated the endogenous immune mechanism of neurons or glial cells and TLR4-associated inflammatory response. In this study, HMGB1 remained expressed in the nucleus of the peri-FCD brain tissue, suggesting that TLR4 may have been upregulated by the release of extracellular HMGB1 in chronic injury caused by repeated epileptic activities in FCD lesion. Furthermore, we also found that the transcript levels of HMGB1 had no significant difference between FCD and peri-FCD tissue. Combined with the results that the protein expression of intracellular HMGB1 in lesion were higher than those in peri-FCD but the total one had no change in lesion, these suggested that repeated epileptic electrical activities had an effect on the cytosolic translocation of HMGB1, but no express levels.
The upregulation of HMGB1-TLR4 pathway in epileptic lesions was likely to be the result of repeated epileptic electrical activities, but whether the upregulation of the pathway will further promote epileptogenesis in turn? No direct evidence on the vicious circle between epileptogenesis and inflammation was reported. In patients with tuberous sclerosis, Zurolo et al. [
9] found that TLR4 and HMGB1 were expressed in the giant embryonic cells of cortical tubers. This finding suggested that the activation of HMGB1-TLR4 pathway may occur earlier than epileptic episodes, which mean that the induction of these inflammatory signaling pathways may be inherent in the occurrence and development of the disease, and its continued activation may be one of the reasons causing the occurrence of late epilepsy. Maroso et al. [
8] found that, in the acute and chronic epilepsy model of adult mice, injecting HMGB1 into the hippocampus of mice significantly increased the duration of seizures and shortened seizure latency, whereas injecting HMGB1 antagonist (BoxA) or TLR4 antagonist (Lps-Rs and Cyp) significantly inhibited the occurrence of acute and chronic seizures. This finding suggested that the upregulation of HMGB1-TLR4 pathway likely promotes the occurrence of epilepsy. In addition, the upregulation of HMGB1-TLR4 pathway occurred in the astrocytes of the FCD brain tissue. In recent years, the role of astrocytes in the occurrence and development of epilepsy has been studied [
15,
16]. Astrocytes, in addition to being an important regulator of the brain’s microenvironment, have been shown to possess the function of innate immune cells. Reactive astrocytes are the source and target of multiple inflammatory factors, such as IL-1β, IL-6, and TNF-α [
17]. In 2017, Shen et al. [
18] found that the postnatal activation of TLR4 in the astrocytes of mice can promote excitatory synaptogenesis in hippocampal neurons. Therefore, we hypothesized that the upregulation of HMGB1-TLR4 and its downstream pathway not only might be the result of repeated epileptic discharge but also promote the occurrence of epilepsy, that is, inflammation and epilepsy were in reciprocal causation, leading to a vicious circle.
Furthermore, we also found that phosphorylation of IKK, IκB-α, and NF-κB as well as expression of two pro-inflammatory cytokines IL-1β and TNF-α in FCD lesion tissue were significantly higher than those in peri-FCD tissue, which suggested that NF-κB inflammatory pathway was activated in FCD lesion. Based on previous studies, the classical NF-κB activation pathway can be mediated through activation of a variety of cell surface receptors, including Toll-like receptors, IL-1 receptor, and TNF receptor, in response to pro-inflammatory mediators like LPS, IL-1, and TNF [
19]. And activated NF-κB translocate from cytoplasm to nucleus and promote the transcription of pro-inflammatory cytokines like IL-1β and TNF-α, thus form the positive feedback [
19]. So, in this study, activated NF-κB inflammatory pathway in FCD lesion was not only caused by upregulation of HMGB1-TLR4 pathway, but could also be a result of cytokine receptor activation.
In our study, we did not found significant association between the level of TLR4 upregulation in FCD lesions and disease duration of epilepsy or seizure frequency. Boer et al. [
20] analyzed a possible correlation between the number of microglia (HLA-DR-positive) and macrophage (CD68-positive) within the dysplastic area and different clinical variables. The result showed that the number of microglia in FCD had a significant correlation with the duration of epilepsy and preoperative seizure frequency. This study suggested that the level of inflammatory response was likely related to clinical variables of epilepsy. We considered that the reason we did not find the correlation in our study is that TLR4 expression in our study was performed in a cross-sectional time, but the level of HMGB1-TLR4 pathway expression might be dynamically changed in the disease duration of epilepsy. Recent findings have shown dynamic changes in HMGB1 and its isoforms in the brain and blood of animals exposed to acute brain injuries and undergoing epileptogenesis [
21].