Inhibition of the NLRP3 inflammasome provides neuroprotection in rats following amygdala kindling-induced status epilepticus

To avoid the interference of estrogen on microglial activation, neuroinflammation and cognitive function [23], only male rats were used in this study. Adult male Sprague-Dawley (SD) rats weighing 260 to 300 g were obtained from the Experimental Animal Center of Qingdao University. The rats were specific-pathogen free rats, regularly checked to ensure the status, and housed in a pathogen-free room with a 12-hour light/dark cycle and given free access to food and water. All experiments were performed in strict accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals. Animal care and sacrifice were conducted according to methods approved by the Qingdao University Animal Experimentation Committee, and all efforts were made to minimize the number of animals used and their suffering. Experimental group allocation and experimental design are shown in Additional file 1: Figure S1.

The NLRP3 siRNA was designed and synthesized by Invitrogen (Carlsbad, CA, USA). Caspase-1 siRNA and control siRNA were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Entranster™-in vivo transfection reagent was purchased from Engreen Biosystem Co, Ltd (Beijing, China). We prepared the Entranster™-in vivo-siRNA mixture according to the manufacturer’s instructions. Briefly, 50 μg siRNA were resuspended in 50 μL RNase-free water to make a siRNA solution. Then, 50 μL of siRNA solution were mixed with 50 μL of Entranster™-in vivo transfection reagent and 100 μL artificial cerebrospinal fluid (aCSF, composition in mmol/L: NaCl 130, KCl 2.99, CaCl₂ 0.98, MgCl₂•6H₂O 0.80, NaHCO₃ 25, Na₂HPO₄•12H₂O 0.039, NaH₂PO₄•2H₂O 0.46) to obtain a 200 μL Entranster™-in vivo transfection mixture. Then, the 200 μL Entranster™-in vivo transfection mixture (containing NLRP3 or caspase-1 siRNA, control siRNA, no siRNA (aCSF only)) was filled into an osmotic pump (Model 2006; ALZET, Cupertino, CA, USA). Meanwhile, rats were anesthetized with 10% chloral hydrate (0.3 mL/100 g, intraperitoneal) and were placed in a stereotaxic apparatus (Stoelting, Wood Dale, IL, USA). A brain-infusion cannula (Brain Infusion Kit 2; ALZET, Cupertino, CA, USA) coupled via vinyl tubing to the osmotic pump was implanted into the dorsal third ventricle (AP: −1.8 mm; L: −0 mm; V: −5 mm). The osmotic pump was placed subcutaneously between the rat scapulae, and the siRNAs were continuously infused into the brain over a 6-week duration at a flow rate of 0.15 μL/hour. This dose of NLRP3 siRNA or caspase-1 siRNA infusion was well tolerated, and no signs of neurotoxicity including hind-limb paralysis, vocalization, food intake, or neuroanatomical damage were observed in preliminary study. Incidentally, the behavioral and biochemical data between control siRNA-treated and non-siRNA-treated rats do not differ. All of the operations were carried out with aseptic techniques.

SE was triggered following the methods of our previous work [24]. Briefly, rats were fixed in a stereotactic apparatus (Stoelting, Wood Dale, IL, USA) under deep anesthesia (10% chloral hydrate, 0.3 mL/100 g, intraperitoneal). The electrodes were permanently implanted into the right basolateral amygdala (AP: −3.0 mm; L: −4.8 mm; V: −8.8 mm) and were connected to a miniature receptacle, embedding in the skull with screws and dental acrylic cement. After electrode implantation, the animals were allowed to recover from surgery for 2 weeks. SE was induced by continuous delivery of 100-ms trains, consisting of 60 Hz 400 μA (peak-to-peak) bipolar 1-ms square-wave pulses, delivered at 60 Hz every 0.5 seconds using a ML1101 electronic stimulator (Nihon Kohden) via the implanted electrode for up to 20 minutes. Electroencephalograms (EEGs) of the right amygdala were recorded with a digital amplifier (AD Instrument, Bio Amp (Shanghai, China), USA). After 20 minutes of continuous stimulation the stimulation was interrupted and the behavioral and electrographic activity of the animals was observed for 60 seconds. If the behavior of the animals indicated the presence of epileptic activity (head nodding or limb clonus), observation was continued for another 5 minutes. If an animal did not meet the criteria of clonic SE (continuous EEG epileptiform spiking and recurrent clonic seizures), stimulation was resumed and the behavior of the animal was checked again after 5 minutes. Once the criteria of SE were achieved, no further stimulation was given. Stimulation period never exceeded 40 minutes. All of the operations were carried out with aseptic techniques. Sham rats were handled in the same manner but without receiving any electrical stimulation. Moreover, our preliminary experiments also show that this amygdala stimulation model is effective.

SRS occurrence was evaluated through a 4-week video monitoring started 2 weeks after SE. All recordings for SRS were done during the light period. Epileptic rats were video-recorded for at least 12 hours daily. The frequency and duration of stage 4/5 seizures were recorded, and the severity of SRS was scored according to Racine’s scale [25]. The recordings were analyzed by observers who were blind to the results of group allocation.

Rats were sacrificed under deep anesthesia and were handled as follows: For Western blot analysis, quantitative real-time PCR, and ELISA, rats were perfused transcardially with 0.9% saline (pH 7.4) only. The brains were removed rapidly and stored in liquid nitrogen until use. For cresyl violet staining and terminal deoxynucleotidyl transferase-mediated dUTP end-labeling assay (TUNEL) analysis, rats were perfused transcardially with 0.9% saline (pH 7.4), followed by a fixative solution containing 4% paraformaldehyde in 0.9% saline (pH 7.4). The brains were removed and fixed in the same fixative at 4°C until use. For double immunofluorescence staining, rat brain was removed without perfusion, embedded in tissue freezing medium, and immediately frozen at −4°C. Frozen tissue was stored at −80°C until sectioning.

Tissues samples were digested with radio immunoprecipitation assay (RIPA) lysis buffer (50 mmol/L Tris-HCl, 150 mmol/L NaCl, 1% Nonidet-40, 0.5% sodium deoxycholate, 1 mmol/L EDTA, 1 mmol/L PMSF) with protease inhibitors (pepstatin 1 μg/mL, aprotinin 1 μg/mL, leupeptin 1 μg/mL) for 30 minutes and centrifuged at 12,000 × g for 15 minutes at 4°C. The protein concentration was determined using the Bradford assay kit (Bio-Rad Laboratories, Hercules, CA, USA). The equal amount of protein from different samples was separated using 8 to12% sodium dodecyl sulfate (SDS) polyacrylamide gels and transferred to polyvinylidene fluoride (PVDF) membranes. The membranes were blocked with 10% non-fat milk in Tween-TBS (TBST) and incubated at 4°C overnight, with the primary antibodies against NLRP3 (1:200; Santa Cruz Biotechnology, Santa Cruz, CA, USA), cleaved IL-1β (1:200; Santa Cruz Biotechnology, Santa Cruz, CA, USA), cleaved caspase-1 (1:200; Biorbyt, San Francisco, CA, USA), and β-actin (1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA, USA). After rinsing, the membranes were appropriately incubated with horseradish peroxidase (HRP)-conjugated suitable secondary antibodies (1:5,000; Zhongshan Inc., Beijing, China) for 2 hours at room temperature. Cross-reactivity was visualized using electrochemiluminescence (ECL) Western blotting detection reagents and analyzed by scanning densitometry using a UVP BioDoc-It Imaging System (UVP, Upland, CA, USA).

Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA), using the protocol supplied by the manufacturer. The cDNA was synthesized using Reverse Transcription System (Bio-Rad, Hercules, CA, USA). The reaction was performed at 42 °C for 50 minutes, 95°C for 5 minutes, and 5°C for 5 minutes, then the cDNA was stored at −20°C. Amplification was carried out using the Stratagene Mx3000P real-time PCR system (Stratagene, La Jolla, CA, USA) and real-time SYBR Green PCR technology (Takara Bio, Inc., Shiga, Japan). Reverse transcription was performed in a final volume of 20 μL containing 2 μL cDNA, 10 μL SYBR Green, 0.4 μL ROX Reference Dye, 0.4 μL forward and reverse primer (1 mol/L), and 6.8 μL nuclease-free water. The optimal conditions were 40 cycles of 95°C for 30 seconds, 60°C for 32 seconds, and 72°C for 30 seconds. All reactions were run in duplicates and the mean values are used. Total RNA concentrations from each sample were normalized by quantity of β-actin mRNA, and the target genes’ expression was evaluated by ratio of the number of target mRNA to β-actin mRNA. Relative expression of genes was obtained by the 2^-△△CT method. Primers were purchased from Invitrogen (Carlsbad, CA, USA) as follows (name: forward primer, reverse primer): nlrp3: 5’-ccagggctctgttcattg-3’, 5’-ccttggctttcacttcg-3’; caspase-1: 5’-aggagggaatatgtggg-3’, 5’-aaccttgggcttgtctt-3’; β-actin: 5’-agggaaatcgtgcgtgac3’, 5’-cgctcattgccgatagtg-3’.

The cytokine analysis was performed in duplicate using commercial ELISA assay kits according to the manufacturers’ instructions. IL-1β was measured in ELISAs from R&D Systems (Minneapolis, MN, USA). IL-18 was measured in ELISAs from Invitrogen (Carlsbad, CA, USA). The results are expressed in pg/mL.

The brains were embedded in paraffin and cut into 7-μm sections. Nissl staining was employed to detect surviving neurons. Briefly, the paraffin-embedded sections were dewaxed and rehydrated according to the standard protocols, and immersed in 1% cresyl violet at 50°C for 5 minutes. After being rinsed with water, the sections were dehydrated in increasing concentrations of ethanol, mounted on the slides, and examined with a light microscope. Only the neurons with a violet nucleus and intact morphology were counted as surviving neurons. The TUNEL staining, which detects DNA fragmentation resulting from apoptotic signaling cascades, was performed to label apoptotic neurons. It may also label cells that have suffered severe DNA damage. Therefore, the TUNEL assay is helpful in identifying seizure-induced neuronal loss in our experiment. We employed the TUNEL assay via a commercial kit according to the manufacturer's instructions (Roche Co., Mannheim, Germany). TUNEL-positive neurons with condensed nuclei were identified as dead neurons. Cell counting was performed on six randomly selected non-overlapping fields in the CA1 and CA3 regions of the hippocampus per slide. The densities of surviving neurons or TUNEL-positive neurons in the hippocampus of the scanned digital images were calculated using Image-Pro Express software (Media Cybernetics, Silver Spring, MD, USA). The total cell counts were averaged from six sections per animal. The survival index was defined as:

Furthermore, the TUNEL-positive neuron index was defined as:

In brief, frozen tissue sections of hippocampus area (20 μm thick) from epileptic rats at 12 hours following amygdala stimulation and time-matched shams were used for double staining of NLRP3/ionized Ca⁺ binding adaptor molecule 1 (Iba1). The sections were obtained by cryosectioning at −20°C, mounted on a glass slide, and incubated at room temperature for 1 hour. Afterward, the sections were fixed in ice-cold acetone for 10 minutes and then dried on a heater for 10 minutes at 40°C. The sections were then blocked with 5% BSA and 0.1% Triton X-100 for 2 hours at room temperature. After a single wash with PBS, sections were incubated overnight at 4°C with a goat polyclonal antibody against Iba1 (1:100, Abcam, Cambridge, UK) as well as a rabbit polyclonal antibody against NLRP3 (1:50; Santa Cruz Biotechnology, Santa Cruz, CA, USA)). Sections were rinsed in PBS and washed 3 times, and then incubated respectively with tetramethylrhodamine (TRITC)-conjugated anti-goat IgG (1:200; Zhongshan Inc., Beijing, China) and fluorescein isothiocyanate (FITC)-conjugated anti-rabbit IgG (1:200; Santa Cruz Biotechnology) for 2 hours at room temperature in a dark and humidified container. After that, the sections were washed with PBS and sealed with a coverslip. The slides were analyzed with a fluorescence microscopy (Olympus, Tokyo, Japan). For nuclear staining we used 4',6-diamidino-2-phenylindole (DAPI). To ensure the specificity of the immunoblotting procedure, control experiments were performed in which the corresponding primary antibody was omitted. Under these conditions, no signal was observed.

Statistical analysis was carried out by SPSS software 17.0 (IBM Inc., Chicago, IL, USA). After confirming normal distribution with skewness and kurtosis statistic test, independent sample t-test or one-way analysis of variance (ANOVA) followed by the Bonferroni post hoc test used to analyze differences among groups. All data were presented as mean ± standard deviation (SD). P < 0.05 was considered statistically significant.

We first investigated whether the expression of IL-1β was altered in rat brain after SE which was induced by amygdala stimulation. Total proteins were extracted from the hippocampal regions of SE rats at different time points (including 3, 6, 12, 24 hours following SE) and of shams, and subjected to Western blot analysis. Because there was no significant difference in all detected variables among different sham groups according to our preliminary analysis, the samples of sham group were collected at 12 hours after the sham operation of amygdala stimulation (without any electrical stimulation). We found that the cleaved IL-1β levels of SE rats were significantly elevated compared to control samples during all time points studied (Figure 1A). In the SE group, cleaved IL-1β levels dramatically increased at 3 hours and reached a maximum at 12 hours (about 4 × those of the sham group ) after SE (P < 0.05; n = 6). Following this maximum, no statistically significant difference was observed in expression level of cleaved IL-1β among rat brains at 24 hours compared with that at 12 hours (P = 1.000; n = 6). Next, we analyzed the expression of hippocampal NLRP3 inflammasome components in SE rats. Importantly, the levels of hippocampal NLRP3 protein, and caspase-1 p20, the products of NLRP3 inflammasome activation, were significantly increased at 3 hours and reached a maximum at 12 hours after SE (both P < 0.05; n = 6, Figure 1B, C). Accordingly, no significant differences were observed in NLRP3 levels (P = 1.000; n = 6), and cleaved caspase-1 levels (P = 1.000; n = 6) among rat brains at 24 hours compared with that at 12 hours after SE. Using double immunofluorescence staining to colocalize NLRP3 with microglia marker Iba1, our result further demonstrated the increased expression of NLRP3 in the Iba1-positive microglia of hippocampus at 12 hours following SE (Figure 1D).

To directly study the potential impact of increased NLRP3 expression levels on the maturation and secretion of IL-1β in SE rats, we knocked down brain NLRP3 expression in animals before undergoing SE by using in-vivo nonviral RNA interference methodology [26]. To evaluate the silencing efficiency of siRNA infusion, the gene expression and protein level of NLRP3 protein were detected by quantitative real-time PCR and Western blotting, respectively. This approach of using NLRP3 siRNA infusion could produce a significant down-regulation of NLRP3 mRNA (by 60%) and protein levels (by 51%) in the brain compared with control siRNA under epileptic conditions (all P < 0.05; Additional file 2: Figure S2 A, B). Incidentally, the NLRP3 levels and gene expression between control-siRNA treated and non-siRNA-treated tissues under epileptic conditions or under control conditions do not differ (all P > 0.05; n = 6, Additional file 2: Figure S2 A, B, and E), excluding potential indirect siRNA-effects on gene expression or the protein level of NLRP3.

Next, we determined the effects of NLRP3 knock down on SE-induced neuroinflammation. The proinflammatory cytokines expression levels were then assessed by Western blot analysis and ELISA assay. Compared to the control-siRNA group, the elevated expression level of cleaved IL-1β in the brain tissues of SE rats could be inhibited by NLRP3 siRNA treatment (P < 0.05; n = 6; Figure 2A). In addition, SE rats infused with NLRP3 siRNA showed a dramatic reduction in brain IL-18 protein levels (P < 0.05; n = 6; Figure 2B). We further examined the changes in caspase-1, which has been known to play a central role in the cleavage of IL-1β and IL-18. The gene expression and protein level of caspase-1 protein were detected by quantitative real-time PCR and Western blotting, respectively. As indicated in Figure 2C, the NLRP3 siRNA treatment could reduce the active caspase-1 protein levels and gene expression in SE rats (both P < 0.05; n = 6).

In above work, we demonstrated that NLRP3 mediated IL-1β over-expression and inflammatory signal activation in the SE rat hippocampus. Accumulating data suggest that inflammation may contribute to epileptogenesis in experimental models as well as in humans. However, whether anti-inflammatory treatments can prevent epileptogenesis remains controversial. Here, we examined the anti-epileptogenic effect of NLRP3 inhibition.

To determine the effect of NLRP3 siRNA on the development and severity of SRS in the chronic phase, SRS were observed from a 4-week video recording, which started 2 weeks after SE. The severity of SRS between siRNA (non or control)-treated SE rats and SE rats do not differ (all P > 0.05; n = 18, Additional file 3: Table S1), excluding an effect of siRNA transfection on status severity.

As revealed by Table 1(A), all rats (18/18) in the control siRNA-treated SE group had SRS during the monitoring period, whereas 66% (12/18) of the rats in the NLRP3 siRNA-treated SE group had SRS. The time to development of SRS was longer in the NLRP3 siRNA-treated SE group compared with the control siRNA-treated SE group (SE + control siRNA group: 14.8 ± 5.5 days; SE + NLRP3 siRNA group: 27.5 ± 9.3 days, Table 1A). The mean number of seizures (total number of seizures/number of recording days, calculated separately for each animal) in the control siRNA-treated SE group and NLRP3 siRNA-treated SE group was 7.58 ± 1.09 and 1.34 ± 0.44 seizures/day respectively. NLRP3 siRNA treatment also dramatically reduced the duration of observed seizures during the chronic epileptic phase (control siRNA, 26.62 ± 6.94 seconds; NLRP3 siRNA, 10.48 ± 2.76 seconds, Table 1A). The differences between the two groups were significant (P < 0.05, Table 1A).

NLRP3, nucleotide binding and oligomerization domain-like receptor family pyrin domain-containing 3; SE, status epilepticus; sec, seconds; SRS, spontaneous recurrent seizures.

If left untreated, SE can cause irreversible brain damage. On this basis, we next investigated the effects of NLRP3 inhibition on neuronal loss in the CA1 and CA3 area of the hippocampus at 6 weeks after SE. Nissl staining was firstly used to detect surviving neurons in the hippocampus. As shown by Figure 3A, a significant increase in neuronal survival rate was noted in CA1 and CA3 regions in the hippocampus of SE rats treated with NLRP3 siRNA compared to the control-siRNA group (CA1: 77.34 ± 5.7% versus 59.57 ± 5.8%, P < 0.05; CA3: 83.43 ± 7.3% versus 62.1 ± 6.5%, P < 0.05). The TUNEL staining assay was then used. As indicated by Figure 3B, a dramatic reduction of the TUNEL-positive neurons index was observed in CA1 and CA3 regions of the hippocampus of SE rats treated with NLRP3 siRNA compared to the control-siRNA group (CA1: 11.41 ± 2.4% versus 15.95 ± 3.3%, P < 0.05; CA3: 9.9 ± 1.9% versus 13.29 ± 2.8%, P < 0.05).

Considering the central role of caspase-1 in the process of IL-1β maturation and secretion, we also successfully knocked down brain caspase-1 in SE rat by in-vivo nonviral RNA interference methodology for 6 weeks (Additional file 2: Figure S2 C, D, and F). We found a significant reduction of IL-1β levels after infused with caspase-1 siRNA (P < 0.05; n = 6, Figure 4A). Furthermore, caspase-1 silencing attenuated the development and severity of SRS in the chronic phase (P < 0.05; n = 18, Table 1B). A marked reduction of TUNEL-positive cell densities in the CA1 and CA3 region of the hippocampus in the SE rat treated with caspase-1 siRNA was also observed (P < 0.05; n = 6, Figure 4B).