Endogenous hydrogen sulphide attenuates NLRP3 inflammasome-mediated neuroinflammation by suppressing the P2X7 receptor after intracerebral haemorrhage in rats

A total of 299 Sprague–Dawley (SD) male rats weighing 280–320 g and 40 1-day-old post-natal Sprague–Dawley male rats were provided by the Experimental Animal Centre of the Third Military Medical University (Chongqing, China). All experimental procedures and animal care procedures were approved by the Ethics Committee of Southwest Hospital, performed in accordance with the guidelines by the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and reported following the ARRIVE guidelines (Animal Research: Reporting in Vivo Experiments, https://​www.​nc3rs.​org.​uk/​arrive-guidelines). The rats were housed in a temperature- and humidity-controlled environment, were provided food and water ad libitum, were maintained under a 12-h light/dark cycle and were acclimatized for more than 1 week before undergoing surgery.

In the present study, the following four separate experiments were conducted:

S-adenosyl-l-methionine, a CBS-specific agonist (SAM, Sigma-Aldrich, St. Louis, MO, USA), and sodium hydrosulfide, a classical exogenous H₂S donor (NaHS, Aladdin, Shanghai, China), were diluted to concentrations of 100 mg/kg [14, 15] and 5.6 mg/kg [7], respectively, in vehicle (saline) solution. The rats were treated intraperitoneally with SAM, NaHS or vehicle at 0.5 h after ICH. For the 72-h study, SAM, NaHS and vehicle were administered three times at 0.5, 24, and 48 h after ICH.

For in vivo adenovirus administration, the following two types of recombinant adenoviruses were used for gene transfer: (1) a replication-deficient human adenovirus containing the rat P2X7R (adenovirus P2X7R, Ad-P2X7R), which was used to upregulate P2X7R expression, and (2) an adenovirus containing human GFP (Ad-GFP), which served as a control for Ad-P2X7R. Ad-P2X7R (6 × 10¹⁰ pfu/mL) and Ad-GFP (2 × 10¹⁰ pfu/mL) were produced by Genechem in Shanghai, China. Both adenoviruses were stored at − 80 °C until use and were diluted to 1.3 × 10¹⁰ pfu/mL in an enhanced transfection solution (Genechem, Shanghai, China) before intracerebroventricular injection. A total volume of 10 μL was used for each adenovirus administration to the animals.

The adenoviruses were injected into the left lateral ventricle 6 days before ICH, according to the previous procedure [16]. The rats were anaesthetised with intraperitoneal injections of sodium pentobarbital (100 mg/kg, Sigma-Aldrich, St. Louis, MO, USA) and then fixed onto a stereotaxic head apparatus. A 26-gauge needle in a 10-μL Hamilton syringe (Microliter 701; Hamilton Company, Reno, NV, USA) was inserted into the left lateral ventricle through a cranial burr hole at the following coordinates relative to the bregma: 1.5 mm posterior, 3.0 mm lateral and 5.5 mm below the horizontal plane of the bregma. Then, the abovementioned formulated solution was slowly injected by a microinfusion pump (Harvard Apparatus, Holliston, MA, USA) at a rate of 0.5 μL/min. The needle was kept in place for 10 min after injection before being slowly withdrawn, and then, the head of the rat was tilted nose-down at a 30° angle for 30 min after suture placement. The animals in the sham and vehicle groups were injected with equal volumes of saline into their left lateral ventricle, and then, their incisions were closed with sutures.

The experimental ICH model was established as described in our previous study [17]. Briefly, the rats were anaesthetised and positioned prone in a stereotactic head frame (Kopf Instruments, Tujunga, CA, USA). A scalp incision was made along the midline, and a burr hole (1 mm) was drilled on the right side of the skull at 3 mm lateral to the bregma and 5 mm ventral to the cortical surface. One microliter (0.2 μL/min) of saline containing 0.2 units of bacterial collagenase (type VII; Sigma-Aldrich, St. Louis, MO, USA) was injected stereotaxically into the striatum (6 mm ventral from the skull surface) using a Nanomite Syringe Pump (Harvard Apparatus, Holliston, MA, USA). After injection, the syringe was left in place for 5 min. The needle was slowly removed over an additional 5 min to prevent backflow. The hole was sealed with bone wax, and the wound was sutured closed. In the sham group, the rats were subjected to only the needle insertion procedure described above. The rats were allowed to recover in separate cages, with free access to food and water.

Primary cultured rat microglial cells were prepared as described in our previous study [18], with minor modifications. Briefly, the meninges, choroidal plexus, brainstem and cerebellum were carefully separated from the cerebral hemispheres of 1-day-old post-natal Sprague–Dawley rats. Then, the cortices of the cerebral hemispheres were dissected and digested with 0.25% trypsin. The fragments were washed twice with D-Hank’s solution, and the resuspended cells were then collected and seeded in uncoated culture flasks containing the indicated medium (DMEM/F12 supplemented with 10% foetal bovine serum, 1 × 10⁵ U/L penicillin, 1 × 10⁵ U/L streptomycin sulphate, pH 7.2) at a concentration of 1 × 10⁶ cells/mL. The cells were cultured at 37 °C in humidified 5% CO₂/95% air, and the medium was changed every 4–5 days. Upon reaching confluence (14 days), the microglial cells were separated from the underlying astrocytic monolayer via gentle shaking of the flasks overnight, a separation made possible by the differential adhesive properties of the two cell lines. The floating microglia were subsequently collected and plated on sterile culture dishes at a density of 1 × 10⁵ cells/cm². After 24 h of culture, the microglial cells were ready for use. The cells were verified via immunohistochemical staining with microglia-specific Iba1 antibodies.

To induce inflammasome activation, we plated 1 × 10⁵ cells in a 6-well plate overnight. The medium was changed to serum-free medium the following morning, and then, the cells were treated with 1 μg/mL lipopolysaccharide (LPS, Sigma, St. Louis, MO, USA) with SAM or NaHS for 12 h. We administered dimethylsulfoxide (DMSO, Sigma, St. Louis, MO, USA)-only treatment as a control. Then, the microglial cells were exposed to 5 mM adenosine triphosphate (ATP) for 1 h. The cell extracts and precipitated supernatants were analysed by immunoblotting.

Neurobehavioural function was assessed with the Modified Neurological Severity Scores (mNSSs) at 1 and 3 days after ICH by an investigator who was blinded to the experimental groups. The mNSS (normal score, 0; maximal deficit score, 18) is determined by summing the scores of motor, sensory, reflex and balance assessments [19], with higher scores signifying more severe neurological deficits.

Brain water content was evaluated for the brain edema as previously reported [16]. Briefly, the brain specimens were quickly divided into the ipsilateral cortex (Ipsi-CX), ipsilateral basal ganglia (Ipsi-BG), contralateral cortex (Cont-CX) and contralateral basal ganglia (Cont-BG). Brain water content was calculated as (wet weight − dry weight)/wet weight × 100%.

Methylene blue assay was performed for the H₂S concentration. The rats were deeply anaesthetised, and their brains were removed post-operatively. The ipsilateral striatum was dissected out on an ice-cold operating table and immediately frozen at −80 °C until assayed. The tissues from the striatum of the injured hemisphere were collected for determination of their H₂S concentrations by the spectro-luminosity method. The brain tissues were homogenized in ice-cold 50 mmol/L potassium phosphate buffer, pH 8.0 (12% wt/vol), with a Polytron homogenizer. The tissue homogenates were centrifuged (47,000g, 10 min, 4 °C), and the cell supernatants were collected to evaluate H₂S levels. Blood samples were collected from the abdominal aorta and were immediately centrifuged to obtain plasma to test H₂S levels. The cell supernatants or plasma samples (75 μL) were mixed with 0.25 mL of Zn acetate (1%) and 0.45 mL of water for 10 min at room temperature. Trichloroacetic acid (TCA; 10%, 0.25 mL) was then added to the mixture, which was centrifuged under the indicated conditions (14,000g, 10 min, 4 °C). The clear supernatant was collected and mixed with N,N-dimethyl-p-phenylenediamine sulphate (20 mmol/L; 133 μL) in 7.2 mol/L hydrogen chloride (HCl) and ferric trichloride (FeCl₃; 30 mmol/L, 133 μL) in 1.2 mol/L HCl. After the resulting solution had incubated for 20 min at room temperature, we measured its absorbance at 670 nm with a spectrophotometer. A calibration curve of absorbance versus sulphide concentrations was constructed using defined concentrations of NaHS solution. When NaHS is dissolved in water, HS− is released and forms H₂S with H+. The H₂S concentration was taken as 30% of the NaHS concentration in each calculation.

The rats were perfused transcranially with phosphate-buffered saline under anaesthesia, and their ipsilateral hemispheres were collected for the following haemoglobin assay to indicate the hematoma volume, as previously reported [20]. Briefly, the rat ipsilateral hemispheres were homogenized in 3 mL of distilled water for 60 s. Then, the homogenized specimens were centrifuged at 13,000g for 30 min, after which 400 μL of Drabkin’s reagent (Sigma-Aldrich, St Louis, MO, USA) was added to 100 μL of supernatant and stored at room temperature for 15 min. The haemoglobin absorbance was measured at 540 nm by a spectrophotometer and quantified using a standard curve. The results are presented as microliters of blood to represent the haemoglobin content in the ipsilateral hemispheres.

Perihaematomal neuronal degeneration was examined by Fluoro-Jade C staining (Millipore, Temecula, CA, USA), as described previously [17]. The brain tissue sections were rinsed in basic alcohol for 5 min before being rinsed for 2 min in 70% alcohol. Distilled water was used to remove the alcohol, after which the sections were incubated in 0.06% potassium permanganate (KMnO₄) for 10 min, stained with 0.0001% Fluoro-Jade C in 0.1% acetic for 10 min and rinsed in distilled water for 5 min. After being triple-rinsed with distilled water, the sections were air-dried for 10 min, cleared in xylene and then covered with slips. Four perihaematomal images in each section were captured by a Zeiss microscope (Zeiss AxioCam, Germany), and Fluoro-Jade C-positive neurons were counted by ImageJ (National Institutes of Health, Bethesda, MD, USA).

Primary microglial cells (1 × 10⁵ cells/mL) were seeded into 96-well plates (100 μL/well) and grouped according to the different concentrations of SAM (0, 10, 100, 20, 400 μmol/L) and NaHS (0, 10, 100, 20, 400, 600 μmol/L) with which they were treated under LPS and ATP stimulation. Cell viability was evaluated by CCK-8 assay (Dojindo Laboratories, Kumamoto, Japan). At the end of each time point, the medium in the 96-well culture plates was changed to DMEM/F12 to avoid background interference, and then, CCK-8 (10 μL) was added to each well. Absorbance was measured at 450 nm using a spectrophotometer, and 620 nm was used as a reference wavelength.

Total RNA was extracted from perihaematomal brain tissue specimens or primary microglial cells using TRIzol reagent (Invitrogen, Camarillo, CA, USA). Isolated RNA was reverse-transcribed into complementary DNA (cDNA) using a cDNA synthesis kit (Vazyme, Jiangsu, China) in accordance with the manufacturer’s protocols. qPCR was performed using synthetic primers and SYBR Green (Thermo, Rockford, IL, USA) with an IQ5 Detection System. After incubating at 50 °C for 2 min and 95 °C for 10 min, the samples were subjected to 40 cycles of 95 °C for 15 s and 60 °C for 1 min. GAPDH was used as an endogenous control gene. The sequences of the primers specific for the P2X7R, NLRP3 and GADPH were as follows:

Samples, including perihaematomal brain tissues, cell lysates and medium, were collected and subjected to western blot analysis, as described in our previous studies [16, 17]. The following primary antibodies were used: rabbit polyclonal anti-CBS antibody (1:200, Santa Cruz Biotechnology, Santa Cruz, CA, USA), rabbit polyclonal anti-P2X7R antibody (1:1000, Alomone Labs, Jerusalem, Israel), rabbit polyclonal anti-NLRP3 antibody (1:200, Santa Cruz Biotechnology, Santa Cruz, CA, USA), rabbit polyclonal anti-ASC antibody (1:500, Abcam, Cambridge, MA, USA), mouse monoclonal anti-caspase-1 p20 antibody (1:200, Santa Cruz Biotechnology, CA, USA), rabbit polyclonal anti-IL-1β antibody (1:1000, Millipore, Billerica, MA, USA), rabbit polyclonal anti-myeloperoxidase antibody (anti-MPO, 1:500, Abcam, Cambridge, MA, USA) and goat polyclonal anti-Iba1 antibody (1:600, Abcam, Cambridge, MA, USA). GAPDH (1:1000, Cell Signalling Technology, Danvers, MA, USA) was used as a loading control. The proteins were detected on nitrocellulose membranes with enhanced chemiluminescence reagents (GE Healthcare, Beijing, China), and the blot bands were quantified by densitometry with Image J software (National Institutes of Health, Bethesda, MD, USA). The results are expressed as a relative density ratio, which was normalized to the mean value of the vehicle or control group.

Frozen sections (20-μm thickness) were obtained from each group at 1 day after ICH, for immunofluorescence staining as described previously [17], and washed with a 0.01 M phosphate buffer solution after 30 min of heating at 37 °C. The sections were incubated for 30 min in 5% bovine serum albumin and then incubated at 4 °C overnight with the following primary antibodies: rabbit polyclonal anti-CBS antibody (1:100, Santa Cruz Biotechnology, Santa Cruz, CA, USA), rabbit polyclonal anti-P2X7R antibody (1:500, Alomone Labs, Jerusalem, Israel), goat polyclonal anti-Iba1 antibody (1:300, Abcam, Cambridge, MA, USA), and rabbit polyclonal anti-MPO antibody (1:50, Abcam, Cambridge, MA, USA).

The microglial cultures were fixed for 30 min in 4% paraformaldehyde. The cells were blocked with 1% bovine serum for 60 min and then incubated overnight at 4 °C with the following primary antibodies: goat polyclonal anti-Iba1 antibody (1:300, Abcam, Cambridge, MA, USA) and rabbit polyclonal anti-P2X7R antibody (1:500, Alomone Labs, Jerusalem, Israel). For the double-staining experiments, the cells were incubated with the above primary antibodies overnight at 4 °C before undergoing a separate incubation with Alexa 488 and Alexa 568 secondary fluorescent antibodies (1:400, Invitrogen, Carlsbad, CA, USA) for 60 min at 37 °C. The nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI, Thermo Fisher Scientific, Waltham, MA, USA) for 10 min. The sections and cells were observed by microscopy (Zeiss, AxioCam, Germany) or confocal microscopy (LSM780, Zeiss, Jena, Germany).

TUNEL staining was performed to evaluate the cell apoptosis in perihaematomal brain tissues. At 1 day after ICH, TUNEL staining was performed with an in situ apoptosis detection kit (Roche Molecular Biochemicals, Mannheim, Germany), according to the manufacturer’s instruction. Briefly, selected sections were pretreated with 20 mg/mL proteinase-K in 10 mM Tris-HCl at 37 °C for 15 min and then incubated in 0.3% hydrogen peroxide dissolved in anhydrous methanol for 10 min after being rinsed in phosphate-buffered saline. The sections were then incubated in 0.1% sodium citrate and 0.1% Triton X-100 solution for 2 min at 4 °C. After several washes with phosphate-buffered saline, the sections were incubated with 50 μL of TUNEL reaction mixture with terminal deoxynucleotidyl transferase (TdT) for 60 min at 37 °C under humidified conditions, and the neuronal nuclei were stained with DAPI. Then, the sections were visualized by a confocal microscope. Negative controls were prepared by omitting the TdT enzyme. To evaluate the extent of cell injury, we examined and photographed six microscopic fields in parallel for TUNEL-positive cell counting.

All data are expressed as the mean ± SEM. SPSS 11.5 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. The Mann–Whitney U test was used to compare behaviour and activity scores among the groups. Other data were analysed by one-way ANOVA followed by the Scheffé F test for post hoc analysis. p < 0.05 was considered statistically significant.

To assess the changes in H₂S production after ICH, we measured H₂S concentrations in the brain striatum and blood plasma at 3, 6 and 12 h and 1, 2, 3 and 7 days after ICH (Fig. 1a). We found that H₂S concentrations in the brain striatum and blood plasma were significantly decreased in ICH-treated rats compared with sham-operated rats at all post-ICH time points (p < 0.05 vs. sham) and reached a nadir at 1 and 2 days after ICH, respectively (p < 0.01 vs. sham). The western blotting results indicated that CBS, a key enzyme involved in endogenous H₂S synthesis, exhibited the same trend in its protein levels as H₂S concentrations in the ipsilateral/right hemisphere after ICH (Fig. 1b). Further investigation with double-immunofluorescence staining revealed that CBS was colocalized with perihaematoma microglia in the ipsilateral striatum, as demonstrated by the presence of Iba1 (Fig. 1c). These data demonstrated that microglia produce endogenous H₂S after ICH and that ICH significantly inhibits CBS activity.

Since H₂S production decreased significantly after ICH, we employed an agonist of CBS, SAM or the classical exogenous H₂S donor, NaHS, to assess the fundamental functions of H₂S. Our data indicated that both SAM administration and NaHS administration significantly increased H₂S content in the brain striatum (p < 0.05 vs. vehicle, Fig. 1d) and blood plasma (p < 0.05 vs. vehicle, Fig. 1e) at 1 day after ICH. These data also demonstrated that SAM and NaHS administration significantly increased endogenous and exogenous H2S content, respectively, at 1 day after ICH.

There was no significant difference in haemorrhage volume among the vehicle group, SAM group and NaHS group (p > 0.05 vs. vehicle, Fig. 1f), demonstrating that our ICH rat models were reproducible and consistent and that the artificial interventions used in this study had no influence on haematoma formation or absorption.

Fluoro-Jade C staining demonstrated significantly reduced perihaematomal neuronal cell injury as a result of SAM or NaHS administration at 1 day after ICH (p < 0.05 vs. vehicle, Fig. 2a, b, d). In addition, compared with the sham group, the vehicle group exhibited an increased number of TUNEL-positive cells in the perihaematomal area (p < 0.01, Fig. 2a, c, e). Remarkably, both SAM treatment and NaHS treatment induced a significant reduction in the number of TUNEL-positive cells in the perihaematomal area (p < 0.05 vs. vehicle, Fig. 2a, c, e). SAM and NaHS administration significantly revered the neurological deficits assessed at 1 and 3 days after ICH (p < 0.01 vs. sham, Fig. 2f) at both time points (p < 0.05 vs. vehicle, Fig. 2f).

At 1 day after ICH, brain water content was significantly increased in the ipsilateral basal ganglia (p < 0.01) and ipsilateral cortex (p < 0.05) of the ICH-treated groups compared with the sham group (Fig. 2g). SAM or NaHS treatment significantly reduced brain water content in the ipsilateral basal ganglia (p < 0.05 vs. vehicle, Fig. 2g), but not in the contralateral basal ganglia and ipsilateral cortex (p > 0.05 vs. vehicle, Fig. 2g). Moreover, there were no significant differences in brain water content in the contralateral basal ganglia and ipsilateral cortex in the ICH-treated groups compared to the vehicle group (p > 0.05, Fig. 2g). At 3 days after ICH, brain water content was increased only in the ipsilateral basal ganglia after ICH (p < 0.01 vs. sham, Fig. 2h) and was significantly reduced by both SAM administration and NaHS administration (p < 0.05 vs. vehicle, Fig. 2h).

Since the key enzyme for the endogenous synthesis of H2S, CBS, colocalized with microglia, as illustrated above, we investigated the effects of SAM and NaHS administration on neutrophil infiltration and microglia accumulation. We detected myeloperoxidase (MPO) and Iba1 levels in brain tissue by immunostaining, as well as western blotting, 1 day following ICH. Immunostaining (Fig. 3a) indicated that SAM and NaHS treatment significantly reduced the number of MPO-positive and Iba1-positive cells in the perihaematomal area in the ICH-treated groups compared to the vehicle group (p < 0.05, Fig. 3b, c). Consistently, the western blotting results showed that MPO and Iba1 protein levels in the ipsilateral hemisphere were markedly reduced after SAM and NaHS treatment in the ICH-treated groups compared to the vehicle group (p < 0.05, Fig. 3d, e).

To investigate the effects of SAM and NaHS on NLRP3 inflammasome activation, we first detected NLRP3 messenger RNA (mRNA) levels by qPCR. NLRP3 mRNA levels were apparently increased in the vehicle group at 1 day after ICH (p < 0.01 vs. sham, Fig. 4a). SAM and NaHS treatment significantly reduced NLRP3 mRNA expression in their respective treatment groups compared to the vehicle group (p < 0.05, Fig. 4a). Then, we detected the protein levels of the NLRP3 inflammasome components and mature IL-1β by western blotting. The protein levels of the NLRP3 inflammasome components, including the protein levels of NLRP3, ASC and caspase-1, and the levels of IL-1β production were evidently elevated in the vehicle group at 1 day after ICH (p < 0.01 vs. sham, Fig. 4b–f). SAM and NaHS treatment significantly suppressed NLRP3 inflammasome component expression and the subsequent secretion of mature IL-1β (p < 0.05 vs. vehicle, Fig. 4b–f).

The qPCR results revealed that ICH-induced P2X7R mRNA expression was evidently elevated in the periphery of the haemorrhage at 1 day after injury (p < 0.01 vs. sham, Fig. 5a). Both SAM treatment and NaHS treatment significantly reduced P2X7R mRNA expression in their respective treatment groups compared to the vehicle group (p < 0.05, Fig. 5a). Likewise, P2X7R protein expression was upregulated at 1 day after ICH induction (p < 0.01 vs. sham, Fig. 5b), and SAM and NaHS administration reversed this trend, as demonstrated by western blot analysis (p < 0.05 vs. vehicle, Fig. 5b).

Immunofluorescent characterization of the brain sections at 1 day after ICH and subsequent confocal analyses were performed to evaluate P2X7R and microglia cell colocalization. Few P2X7R-positive microglia (Iba1-positive cells) were detected in the sham group. However, many more activated P2X7R-positive microglia were observed in the ICH-treated groups post-ICH. SAM and NaHS administration inhibited the increases in the numbers of P2X7R-positive microglia that occurred after ICH induction (Fig. 5c).

To verify the involvement of the P2X7R in the suppression of NLRP3 inflammasome activation by H2S in microglia, we cultured primary microglial cells under in vitro conditions characterized by artificial inflammation and P2X7R activation facilitated by combining LPS and ATP in the culture mediums. First, to verify the primary cultured rat microglial cells, we seeded a subset of cells onto glass slides for characterization by immunofluorescence, which showed that the primary microglial cells were Iba1-positive cells (Fig. 6a, b). LPS and ATP stimulation substantially inhibited cell viability, as demonstrated by the CCK-8 assay (p < 0.01 vs. control, Fig. 6c). These cytotoxic effects were attenuated in a concentration-dependent manner by SAM at concentrations ranging from 50 to 400 μM, and the maximum response was achieved at a concentration of 200 μM (p < 0.01 vs. control, Fig. 6c). Similar effects were observed with NaHS administration. NaHS exerted its maximal effect at a concentration of 400 μM (p < 0.01 vs. control, Fig. 6d). These data suggested that H₂S protected the microglial cells against LPS and ATP stimulation-induced cell injury, results consistent with those of our in vivo study, especially the results of our TUNEL staining assay. SAM (at a concentration of 200 μM) or NaHS (at a concentration of 400 μM) was added to produce endogenous and exogenous H2S for the following in vitro study.

Subsequently, qPCR and western blotting were carried out to detect P2X7R mRNA and protein expression. The results of those experiments revealed that both P2X7R mRNA and P2X7R protein levels were evidently elevated after LPS and ATP stimulation (p < 0.01 vs. control, Fig. 7a, b) and that both SAM treatment and NaHS treatment significantly suppressed these upregulations (p < 0.05, Fig. 7a, b). Immunostaining also indicated that the numbers of P2X7R-positive cells were increased and that the cells were visually identical to activated microglia after LPS and ATP stimulation (Fig. 7c). Both SAM treatment and NaHS treatment significantly alleviated these changes (Fig. 7c).

To study the effects of H2S on microglial NLRP3 inflammasome activation in vitro, we applied qPCR to detect NLRP3 mRNA levels in the microglial cell lysates, and then, we carried out western blotting to detect NLRP3 and ASC protein levels in the microglial cell lysates, as well as caspase-1 p20 and mature IL-1β levels in the medium, under LPS and ATP stimulation. LPS and ATP stimulation significantly enhanced NLRP3 mRNA levels (p < 0.01 vs. control, Fig. 8a). Both SAM and NaHS treatment inhibited this enhancement (p < 0.05 vs. LPS + ATP, Fig. 8a). Compared with the control group, LPS and ATP stimulation increased the protein levels of the NLRP3 inflammasome components in the LPS + ATP group, leading to increased levels of IL-1β and caspase-1 release in the medium (p < 0.01, Fig. 6b–f), but SAM and NaHS treatment significantly suppressed these upregulations (p < 0.05, Fig. 8b–f).

To clarify the role of the P2X7R in the activation of the NLRP3 inflammasome and the subsequent processing of IL-1β, we generated recombinant adenoviruses to overexpress the P2X7R in vivo. The qPCR results revealed that the mRNA levels of the P2X7R and NLRP3 were upregulated in the Ad-P2X7R group (p < 0.05 vs. vehicle, Fig. 9a, b), but not in the Ad-GFP group (p > 0.05 vs. vehicle, Fig. 9a, b), at 1 day after ICH. Administration of both SAM and NaHS inhibited the increase in P2X7R expression (p < 0.05 vs. Ad-P2X7R, Fig. 9a, b). Western blotting performed at 1 day after ICH indicated that similar increases in the protein levels of the P2X7R and NLRP3 were induced by P2X7R overexpression and were suppressed by SAM and NaHS administration (Fig. 9c, d). Moreover, the protein levels of mature IL-1β and MPO were significantly elevated by overexpressing P2X7R at 1 day after ICH in the indicated group compared to the vehicle group (p < 0.05, Fig. 9e, f), but SAM or NaHS administration inhibited these increases (p < 0.05 vs. Ad-P2X7R, Fig. 9e, f).

Additionally, regarding the 3-day study of neurological outcomes and brain oedema in the setting of P2X7R overexpression after ICH, we found that neurological deficits (Fig. 10a, b) and brain oedema (Fig. 10c, d) were worse in the Ad-P2X7R group (p < 0.05), but not in the Ad-GFP group (p > 0.05 vs. vehicle, Fig. 10a–d), than in the vehicle group. Neurological deficits and brain oedema were significantly alleviated in the Ad-P2X7R + SAM and Ad-P2X7R + NaHS groups (p < 0.05 vs. Ad-P2X7R, Fig. 10a–d). These results indicated that the P2X7R played a pivotal role in H₂S-mediated anti-NLRP3 inflammasome activation and neuroprotection after ICH.