All procedures involving animals were approved by the Animal Studies Committee of Third Hospital of PLA and Weihai Municipal Hospital and complied with the Guide for the Care and Use of Laboratory Animals (Guide; NRC 2011). In this study, wild type (WT) mice were purchased from the Animal Center of the Third Military University (Chongqing, China), TLR4^−/− were purchased from the Model Animal Research Center of Nanjing University (Nanjing, Jiangsu, China). According to a previously reported method (Chen et al., 2011; Anand et al., 2012), NLRP6^−/− mice were generated by the replacement of exons 1 and 2 of the Nlrp6 gene (N-terminal domain) with the internal ribosome entry site–β-gal–neomycin resistance cassette using a targeting vector (with the help of Beijing Biocytogen Co., Ltd). Eight to ten-week-old or 22 ± 2 g mice were used in study. All mice had a C57BL/6 background, and only male was used. The mice were provided food and water ad libitum, and were housed under a 12 h light and 12 h dark cycle.

The detailed procedures used to construct the ICH model were established in our previous studies (Wang et al., 2013; Xiong et al., 2016a). Mice were briefly anesthetized intraperitoneally with 4% phenobarbital sodium and were immobilized on the stereotaxic apparatus (RWD Life Science Co, Shenzhen, China). All experimental mice received a total of 20 μL autologous blood injected into the caudate nucleus (bregma 0: 0.8 mm anterior, 2 mm left lateral and 3.5 mm deep) successively. The needle of microsyringe was held in place for 10 min after injection to avoid the blood back streaming. The craniotomy was then sealed with bone wax, and the scalp was closed with sutures. Body temperature was maintained at 37°C throughout the procedure. The mice that died due to anesthesia and unsuccessful ICH models, including asymptomatic and dead mice before euthanasia were excluded from this study. About 5%–10% mice were excluded in study because of the death before sacrifice or without symptoms caused by ICH.

TAK242 (Takeda Pharmaceutical Co) was formulated with 1% dimethyl sulfoxide (σ) and double-distilled water to a final concentration of 0.4 mg/mL and then injected intraperitoneally at a dose of 3 mg/kg beginning 1 h after ICH and giving another injection 6 h after the first injection.

At the designated time points (6 h, 12 h, 1 day, 3 days, 5 days and 7 days after ICH or sham group and normal brain), the animals were prepared for analysis. To prepare tissues for western blot and real-time PCR (RT-PCR), perihematomal brain specimens were frozen in liquid nitrogen and stored at −80°C until further processing. For immunohistochemistry, the animals were perfused transcardially with 0.9% NS followed by 4% paraformaldehyde in 0.1 M PBS, and the brains were immersed in sucrose and then embedded in paraffin, cut into 6 mm coronal sections, and mounted on coverslips coated with polyL-lysine. For further immunofluorescence staining, the mouse brains were post-fixed for 24 h and then immersed in 0.1 M PBS containing increasing amounts of sucrose (15%–30%) at 4°C for 12 h. The brains were then cut into 30 μm coronal sections.

According to the manufacturer’s instructions, total RNA was isolated from perihematomal brain tissues with a Classic Total RNA isolation kit (Invitrogen, USA). RNA concentrations were determined photometrically by measuring optical density at 260 nm (OD260). Reverse transcript to obtain cDNA using a PrimeScript™ RT Reagent Kit with gDNA Eraser (Takara, Japan). Real-time PCR was conducted in accordance with the instructions of Maxima SYBR Green/ROX qPCR Master Mix (Fermentas, Canada). The experimental data were analyzed using the 2^−ΔΔCt method. All independent experiments were repeated three times. The mRNA expression of genes of interest was normalized by the expression of GAPDH. The primer sequences for mouse NLRP6 inflammasome (forward: 5′AGCTGAGAACGCTGTGTCG3′; reverse: 5′AACTTGGGAACCCCGAAGC3′), and GAPDH (forward: 5′ACCACAGTCCATGCCATCAC 3′; reverse: 5′ACCTTGCCCACAGCCTTG 3′) were synthesized by Invitrogen.

According to our previous protocol (Wang P. F. et al., 2014; Xiong et al., 2016a), total protein extracts were prepared from perihematomal brain tissues. The proteins were separated on 15% polyacrylamide gels by standard SDS polyacrylamide gel electrophoresis and then transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, USA) using a semidry electroblotting system (Transblot SD, Bio-Rad). The PVDF membranes were blocked with 5% non-fat dry milk in 0.05% Tween-20 in PBS for 1 h at room temperature and were then incubated with a goat polyclonal anti-NLRP6 antibody (1:1000, Santa Cruz, CA, USA), a mouse monoclonal anti-NF-κB p65 (1:1000, Santa Cruz, CA, USA) and a rabbit monoclonal anti-GAPDH antibody (1:1000, Cell Signaling Technology, Danvers, MA, USA) at 4°C overnight. The membranes were further incubated with HRP conjugated goat anti-rabbit secondary antibodies (1:6000, Zhongshan Golden Bridge Inc., China) at 25°C for 1.5 h. Bound antibodies were visualized using a chemiluminescence detection system. The signals were measured by scanning densitometry and computer-assisted image analysis. The optical density of each band was evaluated with ImageJ software. Ratios of NLRP6/GAPDH were calculated from the optical densities.

According to our previous methods (Fang et al., 2014; Xiong et al., 2016a), the paraffin brain sections were deparaffinized by sequential washing with xylene and graded ethanols. Three-percent hydrogen peroxide was used to block the endogenous peroxidase activity for 10 min, and antigen retrieval was conducted in 10 mmol/l citrate buffer (pH 6.0) for 15 min at 100°C. Then, the sections were treated with 5% bovine serum albumin to block nonspecific staining for 30 min. The sections were then incubated overnight at 4°C with the NLRP6 antibody (1:100, Santa Cruz, CA, USA). Following PBS washing, the sections were incubated with a goat anti-rabbit HRP secondary antibody (1:500, Boster, China) for 1 h at 37°C. All immunostaining was performed at the same time under the same conditions. Three sections were prepared for each brain tissue.

The immunofluorescent staining was performed according to our previously reported methods (Xiong et al., 2016a). Twenty-five-micron brain sections were incubated with the following antibodies: NLRP6 (1:100, Santa Cruz, CA, USA), CD11b (1:200, Santa Cruz, CA, USA), NeuN (1:200, Millipore, USA), GFAP (1:100, Sigma, MO, USA) and CD31 (1:200, Santa Cruz, CA, USA). The secondary Abs included Alexa Fluor 647 (1:200, donkey anti-mouse), Alexa Fluor 594 (1:200, donkey anti-rabbit), and Alexa Fluor 488 (1:200, donkey anti-rabbit, donkey anti-mouse, and donkey anti-rat), all from Invitrogen (Carlsbad, CA, USA). The mature slices were imaged by confocal microscopy (Leica TCS Sp5, Mannheim, Germany). Quantification of double-labeled cells was also performed as previously described (Xiong et al., 2016a), and the results were shown as “Number of cells/per region”. Three sections were prepared for each brain tissue specimen; automatic counting was performed in brain tissues for double-labeled cells, and the averaged data obtained by three researchers who were blinded to the group information represented the value of each brain.

Perihematomal brain tissues collected from each group were homogenized and centrifuged, and the supernatant was collected for analysis. The concentrations of inflammatory factors, including IL-1β, IL-6 and tumor necrosis factor-α (TNF)-α were measured using an enzyme-linked immunosorbent assay (ELISA) reagent kit (Dakewe Biotech Company) following the manufacturer’s instructions.

The mice were anesthetized on 1, 3, 5 and 7 days after ICH, and ipsilateral brain tissues were obtained. Then, the brain water content (BMC) was measured according to our previously established methods (Xiong et al., 2016a). Briefly, the obtained brain tissues were dabbed with filter paper to remove surface water and were subsequently evaluated by an electronic balance to determine the wet weight. Then, the brain tissues were dried at 100°C for 24 h to determine the dry weight. The following formula was used to calculate water content: BMC (%) = wet weight − dry weight)/wet weight × 100%.

The Neurological Deficient Score (NDS) were measured on 1, 3, 5 and 7 days after ICH according to our previous reports (Wang et al., 2013; Xiong et al., 2016a). Scoring was performed by three trained investigators who were blind to the group information, and the average values were the final score for each mouse.

All of the data are presented as the mean ± SEM. Analysis was performed using SPSS 13.0 software. The t-test for independent samples was used to compare two groups, and comparisons among multiple groups were examined using one-way analysis of variance (ANOVA), followed by Scheffé post hoc test. Differences were considered significant at a P < 0.05.

Although NLRP6 inflammasome was first studied almost 10 years ago, little is known about NLRP6 inflammasome in the CNS diseases. To help determine the expression of NLRP6 inflammasome in brain after ICH, we first performed NLRP6 inflammasome mRNA expression profiles of perihematomal brain tissues at different time points. We found that NLRP6 inflammasome expression was markedly increased at 1 day after ICH compared with the normal brain (NS) and sham groups, while the upregulation of NLRP6 inflammasome was gradually decreased from 3 days after ICH (Figure 1A). To confirm the RT-PCR results, we next performed western blot to further measure the NLRP6 inflammasome protein level and found that it was gradually increased in the perihematomal brain tissues at the 6 h, with a peak at 1 day after ICH (Figure 1B).

Having characterized the expression profiles of NLRP6 inflammasome in the perihematomal brain tissues of mice caused by ICH, we next explored the cell source of NLRP6 inflammasome. First, the results of immunohistochemical staining showed that NLRP6-positive cells were significantly increased in perihematomal brain tissues at 1 day after ICH, and most of the expression occurred in cells with glial-like morphology, while no expression of NLRP6 inflammasome in the NLRP6^−/− mice at 1 day after ICH (Figure 2). Immunofluorescence staining was performed to further investigate the cell source of upregulated NLRP6 inflammasome after ICH. To enable direct comparison with the data from the experiment described above, we also performed this experiment at 1 day after ICH. GFAP-positive astrocytes appeared to be the major source of NLRP6 inflammasome expression after ICH, in contrast, little NLRP6 inflammasome expression was detected in NeuN-positive neurons or without expression in CD11b-positive microglia and CD31-positive endothelial cell (Figure 3). While the NLRP6^−/− mice showed no NLRP6 inflammasome expression colocalized with GFAP-positive astrocytes after ICH, and little colocalization was detected in the normal brains (Supplementary Figure S1). These results demonstrate that astrocyte-derived NLRP6 inflammasome may be involved in the neuroimmune response to ICH-induced brain injury.

Next, we investigated the roles of NLRP6 inflammasome in brain injury caused by ICH. Brain edema, neuroinflammatory levels and NDS are important indexes for assessing ICH-induced brain injury severity. First, we measured the proinflammatory cytokines, including TNF-α, IL-1β and IL-6, at different time points after ICH using the ELISA kits. The results showed that the levels of IL-1β, IL-6 and TNF-α were significantly increased in the NLRP6^−/− mice compared with the WT mice at 1, 3 and 5 days after ICH (Figure 4A). Next, we observed BMC in mice after NLRP6 deficiency compared with the WT group at all time points using the dry-wet weight method. We found that BMC in the perihematomal brain tissues on day 3 and 5 were significantly increased in NLRP6^−/− mice compared with the WT group after ICH (Figure 4B). Additionally, we found that the NF-κB activity was increased in the NLRP6^−/− mice after ICH when compared to the WT mice (Figure 4C). Furthermore, we investigated the biofunctional role of NLRP6 inflammasome in ICH using the NDS. We found that the NDS scores were significantly increased in NLRP6^−/− group when compared to the WT group (Figure 4D). These results suggested that enhanced expression of NLRP6 inflammasome protects against the ICH-induced brain injury.

Next, we investigated the regulative mechanism of increased NLRP6 inflammasome expression after ICH. As the NLRP6 inflammasome is the intracellular innate immune sensors (Strowig et al., 2012), and the TLR4 is the extracellular innate immune sensors (Akira and Takeda, 2004), which could be triggered by hematoma productions, such as heme, hemoglobin (Hb), etc., after ICH (Lin et al., 2012; Wang Y. C. et al., 2014). Therefore, we want to know whether the extracellular innate immune sensors could influence the intracellular one. Using the TLR4^−/− mice, we found that NLRP6 expression was markedly decreased at 1 day after ICH when compared to the WT mice both at protein (Figures 5A,B) and mRNA levels (Figure 5C). The same results were also obtained by using the TLR4 antagonist TAK242 (Figures 5D,E). Furthermore, the results of immunofluorescent staining also showed that double- (NLRP6 and GFAP) positive cells were lower in TLR4^−/− and TAK242 treated mice than in C57BL/6 and Vehicle treated mice at 12 h after ICH (Figure 6). These results indicated that enhanced NLRP6 inflammasome expression may via the TLR4 signaling after ICH.