BV2, a line of mouse microglial cells, was purchased from Guangzhou Jennio Biotech Co., Ltd (Guangzhou, China) and cultured in DMEM at a density of 2 $\times$ 10${}^5$/mL with 10% fetal bovine serum added as a supplement (Cat. no. 12800017 and 10099141, Invitrogen, Carlsbad, CA, USA) in an incubator for cell culture. Cells were treated with emodin or NLRP3 inhibitor MCC950 (Cat. no. HY-14393 and HY-12815, both from Monmouth Junction, NJ, USA) at doses of 5, 10, 20, or 50 $\mu$M, respectively for 18 h prior to the addition of LPS (1 $\mu$g/mL; Cat. no. L4391, Sigma-Aldrich, St. Louis, MO, USA). ATP (5 mM; Cat. no. A1852, Sigma-Aldrich, St. Louis, MO, USA) was used on the cultures 6 h after LPS application and incubated for 30 min. The treatment-induced morphological changes in BV2 cells were observed with an inverted phase-contrast microscope (Model AxioVert 200M, Zeiss, Oberkochen, Germany) and in order to do further study, the cells and supernatants were then collected.

The cell line of mouse hippocampus neurons HT-22 (Jennio Biotech, Guangzhou, China) was cultured in DMEM in a cell culture incubator. The cell cultures were treated with emodin (20 $\mu$M) for 18 h, then LPS (1 $\mu$g/mL) was added and incubated for another 6 h. The cells were then treated with ATP (5 mM) and cultured for 30 min. In parallel, the upper compartment of a transwell plate included DMEM containing BV2 cells suspended at a density of 5 $\times$ 10${}^5$ cells/mL, the lower compartment was used to hold HT-22 cells and then cultured for 1 h before the addition of emodin. The cocultures were then stimulated with LPS and ATP as above and the lower compartment’s cells and supernatants were gathered for additional examination. The ratio of BV2 cells to HT-22 cells was 1:1.

To detect the toxicity of emodin or LPS and ATP co-stimulation, 5 $\times$ 10${}^3$ cells per well of plated BV2 or HT-22 cells were cultured in 96-well plates with 100 $\mu$L medium overnight. Then, the culture had successive emodin treatments or MCC950, LPS, and ATP at indicated concentrations. 10 $\mu$L of the CCK-8 reagent (Cat. no. CK04, Dojindo, Kumamoto, Japan) was added into each well after 24 h, and they were left to incubate for 2 h in dark. A microplate reader (Model Synergy 2, BioTek, Winooski, VT, USA) was used to measure the absorbance at 450 nm.

Proteins from cultured BV2 cells were extracted with lysis buffer (2% SDS, 62.5 mM Tris-HCl pH 6.8, 10% glycerol, 4% $\beta$-mercaptoethanol, 1 mM phenylmethanesulfonyl and 50 mM DTT). Protein samples (30 $\mu$g/lane) were electroblotted onto polyvinylidene fluoride membranes after being separated by 8% or 12% SDS-PAGE (1658033, Bio-Rad, Hercules, CA, USA). The membranes were treated with primary antibodies against NLRP3, cleaved caspase-1, ASC, or GSDMD-N (all diluted at 1:500; ab263899, ab179515, ab307560, and ab209845, from Abcam, Cambridge, UK) at 4 °C overnight after being blocked with 5% skimmed milk. As an internal standard, anti-$\beta$-actin antibody (1:3000; ab8224, Abcam) was applied. The membranes were scanned using an Odyssey laser scanning system (LI-COR, Lincoln, NE, USA) after incubating with appropriate fluorescence-conjugated secondary antibodies (1:5000; 926-32211 and 925-68070, LI-COR, Lincoln, NE, USA) for 1 h at room temperature. By using Image J software’s (version 2.0, LOCI, University of Wisconsin, Madison, WI, USA) densitometric analysis, the protein bands were measured and normalized to the internal standard $\beta$-actin.

Following the manufacturer’s instructions, the levels of TNF-$\alpha$, IL-1$\beta$, or IL-18 in BV2 or HT-22 cell culture supernatants were determined by ELISA (Cat. no. BMS607-3 and BMS6002, eBioscience, San Diego, CA, USA, for TNF-$\alpha$ and IL-1$\beta$, and Cat. no. YJ003057, Mlbio, Shanghai, China, for IL-18). The absorbance was measured using a microplate reader at 450 nm (Model Synergy 2, BioTek, YT, USA).

4% paraformaldehyde was used to fix the BV2 cells for 20 min at room temperature, followed with incubation in PBS with 0.3% Triton X-100 and 3% normal goat serum for 30 min. Then, cells were incubated with the primary antibody against ASC (1:100; ab307560, Abcam) overnight at 4 °C, followed by incubation with FITC-conjugated secondary antibody (1:400; ab6717, Abcam) for 18 h at 4 °C. To mark the nuclei, DAPI was used as a counterstain on BV2 cells, and the stained cells were quantified in five randomly selected visual areas on each coverslip under fluorescence microscopy (DMLB, Leica, Heidelberg, Germany). The ratio of DAPI-labeled cells to ASC-positive cells in each field of view was calculated for statistical analysis.

After fixation and permeability as above, the HT-22 cells were incubated with primary antibody against NeuN (1:400; ab104224, Abcam) for 24 h at 4 °C. Cells were immediately incubated with Alexa Fluor 594-conjugated secondary antibody (1:400; ab150120, Abcam) for 18 h at 4 °C. To visualize apoptotic cells, the Fluorescein in situ cell death detection kit (Cat. no. 11684795910, Roche, Basel, Switzerland) was used in the TUNEL experiment as directed by the manufacturer. All images were visualized by fluorescence microscopy (DMLB, Leica, Germany) at 200$\times$ magnification. For quantification, TUNEL-positive cells and NeuN-positive cells were counted in five randomly selected visual areas on each coverslip to calculate the proportion of cells with TUNEL staining to the total number of NeuN${}^{+}$ and TUNEL${}^{+}$ cells in a given field of view.

LDH is an enzyme that is normally found in the cytoplasm and is released into the medium when the cell membrane is disrupted, thus, the quantification of LDH release can be used to evaluate the integrity of cell membranes. After LPS and ATP treatment, using an LDH Assay kit (Cat. no. K726, BioVision, Mountain View, CA, USA) in accordance with the manufacturer’s instructions, the cell supernatants were gathered for LDH measurement. LDH oxidizes lactate to produce NADH, reacting with WST, which can be measured at 450 nm optical density using an absorbance plate reader. Therefore, the volume of LDH that cells secreted was measured at 450 nm optical density with an absorbance plate reader (Model Synergy 2, BioTek, VT, USA), and then, the values were interpolated using the calibration curve and the LDH release finally was calculated and expressed as the value of U/L, which represents the degree of cell necrosis.

The vitality of the HT-22 cell was assessed using Syto-13/PI, two dyes with various membrane permeabilities. Syto-13 is used to identify viable cells, whereas a non-permeant membrane red dye called PI is used to spot cells that have changed membrane integrity and are either necrotic or in the advanced stages of apoptosis. Briefly, cells were incubated with 0.1 $\mu$M Syto-13 (Cat. no. S7575, Molecular Probes, Eugene, OR, USA) and 10 $\mu$g/mL PI (Cat. no. P4170, Sigma-Aldrich, MO, USA) in PBS at 37 ºC for 5 min and visualized using a laser confocal microscope (TCS SPE, Leica, Heidelberg, Germany). Both viable cells and necrotic cells in five visual fields on each slide were counted to determine the typical proportion of PI+ necrotic cells.

Data are reported as mean $\pm$ SD, and one-way ANOVA is used to examine them before the Newman-Keuls post hoc test. Statistical analysis was conducted using the Statistics Package for Social Science (SPSS, 16.0, IBM Corp., Chicago, IL, USA) and differences were deemed statistically significant at p 0.05.

Emodin had no conspicuous cytotoxicity to BV2 cells during 24 h treatment at 20 $\mu$M (Fig. 1b), therefore, 20 $\mu$M was selected as the greatest concentration for the studies that followed.

The impact of emodin on LPS/ATP-induced BV2 cell pyroptosis was investigated under the inverted phase-contrast microscope (Fig. 1c). LPS/ATP caused the formation of abnormal cells, such as dysfunction of pores and cellular membrane damage, as well as irregular shrinkage or formation of closed-cell foams or bubbles. Necrosis was considerably increased. Emodin treatment (5, 10, and 20 $\mu$M) reduced these effects and maintened the normal morphology of LPS/ATP-treated BV2 cells in a dose-dependent manner (Fig. 1c). This was accompanied by increased cell viability (Fig. 1d) and alleviated LPS/ATP-induced BV2 membrane damage (Fig. 1e).

IL-18 and IL-1$\beta$, two proinflammatory cytokines, play a vital role in pyroptosis and are released as a consequence of caspase-1 and ‘inflammasome’ activation. TNF-$\alpha$, as an inflammatory factor in the brain, can directly activate microglia to enhance the inflammatory response in the CNS by affecting the BBB permeability, thereby invading the center and causing CNS inflammation. LPS/ATP significantly elevated the level of IL-1$\beta$, IL-18, and TNF-$\alpha$, while emodin treatment (10 and 20 $\mu$M) remarkably suppressed the production of IL-1$\beta$, IL-18, and TNF-$\alpha$ comparison with the control group. Our findings show that emodin suppresses the proinflammatory cytokine production induced by LPS/ATP stimulation in a dose-dependent manner (Fig. 2).

Fig. 2.

Emodin decreased the pro-inflammatory cytokine release in LPS/ATP-induced BV2 cells. BV2 cells were co-stimulated with LPS and ATP after 18 h pretreatment with emodin or MCC950. The TNF-$\alpha$, IL-1$\beta$, and IL-18 was detected by ELISA. Values are represented by the mean $\pm$ SD of five independent experiments. Data were analyzed by one-way ANOVA followed by Newman Keuls post hoc test. ** p 0.01 compared with the control group, + p 0.05, ++ p 0.01 compared with LPS + ATP. group

NLRP3 receptor protein does not possess pro-inflammatory activity. NLRP3 combines with caspase-1 and ASC to form the NLRP3 inflammatory body, which becomes proinflammatory. The activation of the NLRP3 inflammasome was examined using green fluorescence to label ASC and blue fluorescence to locate the nuclei in BV2 cells. Compared to the control group, a large number of NLRP3 inflammasomes were observed in the LPS/ATP stimulated cells, suggesting that LPS/ATP could activate NLRP3 inflammasomes. MCC950, the NLRP3 inflammatory body specific inhibitor, significantly inhibited the NLRP3 inflammatory body activation. Emodin also suppressed the formation of the NLRP3 inflammasome in a dose-dependent manner (Fig. 3a,b), suggesting that emodin can suppress the NLRP3 inflammatory body activation.

Fig. 3.

Emodin suppresses pyroptosis via NLRP3/caspase-1 axis inhibition in BV2 cells. BV2 cells were treated as above. (a) BV2 cells were subjected to immunofluorescent staining for ASC and the nuclei were stained using DAPI. Scale bar = 50 $\mu$m. (b) The ASC-positive cell number was counted and the ratio of ASC-positive cells to DAPI-labeled cells was calculated. (c,d) The protein expression of NLRP3, cleaved caspase-1, ASC, and GSDMD-N were estimated by western blotting. Values are expressed as the mean $\pm$ SD of three or four independent experiments. Data were analyzed by one-way ANOVA followed by Newman Keuls post hoc test. ** p 0.01 compared with the control group, + p 0.05, ++ p 0.01 compared with LPS + ATP group.

Pyroptosis is carried out by the pyroptosis-executing protein GSDMD and is triggered by NLRP3 activation. It was showed that emodin pretreatment significantly restrained the pyroptosis-related protein levels of NLRP3, ASC, caspase-1, and GSDMD-N in LPS/ATP-treated BV2 cells, similar to the effect of MCC950 (Fig. 3c,d). This suggests that emodin can prevent LPS/ATP-caused BV2 cell pyroptosis by inhibiting NLRP3 inflammasome activation.

Under normal conditions, there were no significant differences in LDH, optical density (OD) values measured by CCK8, and apoptotic rates between the HT-22 neuron group and the HT-22 neuron co-cultured group (Fig. 4). After LPS/ATP treatment, the release of LDH was significantly enhanced in the co-culture group (Fig. 4f), neuronal OD values were prominently decreased (Fig. 4e), and the neuronal apoptosis rate was improved (Fig. 4c,d), suggesting that microglial pyroptosis aggravates the degree of neuronal injury under LPS/ATP conditions.

Fig. 4.

Emodin alleviates LPS/ATP-induced neuronal toxicity by inhibiting inflammasome-mediated pyroptosis. LPS and ATP were applied to the HT-22 neuronal mono-cultures or the transwell co-cultures of HT-22 neurons and BV2 microglia after emodin pretreatment. NeuN and TUNEL double-staining was performed in HT-22 cells to assess the neuroprotection influence of emodin. Representative images are demonstrated in (a) and statistical data for the proportion apoptotic cells expressed as the ratio of TUNEL-labeled cells to the sum of NeuN${}^{+}$ and TUNEL${}^{+}$ cells are shown in (c). (b) The HT-22 cells were double stainned with SYTO-13/PI. (d) The percentage of PI-labeled cells in the sum of SYTO-13${}^{+}$ and PI${}^{+}$ cells was reported in the histogram. Scale bar = 50 $\mu$m. The HT-22 cell viability was detected using the CCK-8 (e) and LDH assays (f). All the data were analyzed by one-way ANOVA followed by Newman Keuls post hoc test. ** p 0.01 compared with the control group, ++ p 0.01 compared with LPS + ATP group in HT-22 mono-cultures, && p 0.01 compared with LPS + ATP group in HT-22 and BV2 co-cultures.

The supernatants obtained from the co-culture models were analyzed to confirm the neuronal damage caused by excessive inflammatory mediator release, showing that LPS/ATP stimulation resulted in markedly increased concentrations of TNF-$\alpha$, IL-1$\beta$, and IL-18 (Fig. 5). Since emodin can inhibit neuroinflammation associated with microglial pyroptosis in BV2, we investigated whether emodin could reduce microglial pyroptosis-mediated neuronal injury. Emodin added to the culture medium of LPS/ATP-treated HT-22 neurons co-cultured with BV2 cells significantly decreased LDH levels (Fig. 4f), increased neuronal viability (Fig. 4e), and significantly decreased TNF-$\alpha$, IL-1$\beta$, and IL-18 (Fig. 5). The TUNEL assay and Syto-13/PI assay revealed a reduced apoptosis rate of neurons (Fig. 4a–d). Taken together, these findings imply that emodin can protect neurons via suppressing LPS/ATP-induced microglial pyroptosis.

Fig. 5.

Emodin is neuroprotective by inhibiting inflammatory mediator release from microglia. Emodin was applied to the transwell co-cultures of HT-22 neurons with BV2 microglia, followed by LPS and ATP co-stimulation. Cell culture supernatants in lower compartment were harvested to determine the concentration of TNF-$\alpha$, IL-18, and IL-1$\beta$. All the data in the histograms were from four independent experiments and analyzed by one-way ANOVA followed by Newman Keuls post hoc test. ** p 0.01 compared with the control group, ++ p 0.01 compared with LPS + ATP group.