Adult male Sprague-Dawley (SD) rats, with an average weight of 200 $\pm$ 20 g, were procured from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). The rats were housed under controlled conditions, which included a 12-hour light/dark cycle (with illumination commencing at 07:00), an ambient temperature maintained at 21 $\pm$ 2 °C, and a relative humidity of 50 $\pm$ 5%. All subjects were provided with ad libitum access to food and water.

In Experiment 1, rats were distributed according to chance into six groups: control; 1 day of restraint stress combined with ice-water swimming (1 d RS + IS); 3 d RS + IS; 7 d RS + IS; 14 d RS + IS; and 21 d RS + IS (n = 10).

In Experiment 2, rats were distributed according to chance into four groups: control, 21 days of restraint stress combined with ice-water swimming (stress), stress + Necrosulfonamide (stress + NSA, NSA: a GSDMD inhibitor), and NSA (n = 10).

In Experiment 3, rats were distributed according to chance into six groups: control (n = 12), 21 days of restraint stress combined with ice-water swimming (stress, n = 12), stress + GC (1 mg/kg) (n = 12), stress + GC (2 mg/kg) (n = 6), stress + GC (3 mg/kg) (n = 6), and GC (1 mg/kg) (n = 6). All the experimental procedures are shown in Fig. 1.

Fig. 1.

Overall experimental flowchart. ELISA, enzyme-linked immunosorbent assay; LDH, Lactate dehydrogenase; HE Staining, Hematoxylin & eosin (HE) staining.

The stress treatment was carried out as previously described [14]. The rats were confined and couldn’t move around freely (8:00 to 14:00) without food or water. Rats were then forced to swim for 5 min every day in frigid water. The duration of stress treatment was 1, 3, 7, 14, and 21 days. At the same time, control rats were deprived of food and water and kept in cages. The body weights of all rats were measured at 1, 3, 7, 14, and 21 days before modeling. We selected the 21-day stress group with notable pathological and behavioral alterations for mechanism study in follow-up experiment 2 and experiment 3.

In experiment 2, after 21 days of stress, the rats in the stress + NSA and NSA groups were injected intraperitoneally with 10 mg/kg of NSA (MCE, HY-100573, Monmouth Junction, NJ, USA; dissolved in 20% dimethyl sulfoxide (DMSO, MCE, HY-Y0320C)) at 08:00 for 7 days.

In experiment 3, after 21 days of stress, the rats in the stress + GC and GC groups were injected intraperitoneally with different concentrations of GC (Dexamethasone, a long-acting glucocorticoid, MCE, HY-14648, Monmouth Junction, NJ, USA; dissolved in normal saline) at 08:00 for 7 days.

Open Field Test was conducted with SuperMaze (Shanghai, China). Rats were put into a square arena (0.9 m $\times$ 0.9 m $\times$ 0.3 m) that was split into sixteen squares of the same size on the second day after the stress treatment. Rats were let to explore freely for 5 min in the open field. A video camera was used to record the rats’ movements. The core zone was defined as the four center squares, while the remaining squares were classified as the peripheral region. Following the testing, the quantity of rat feces was counted.

Tissue handling followed established protocols [14]. On the second day after the open field test, 1% pentobarbital sodium (50 mg/kg, P3761, Sigma-Aldrich, St. Louis, MO, USA, in the vehicle: 0.9% saline) was administered intraperitoneally to anesthetize the rats. Then blood was extracted from the rats’ abdominal aorta while rats were anesthetized. Subsequently, rats were executed by beheading. The brains were taken out and immediately preserved in 10% formalin (Sangon Biotech, A375256, Shanghai, China, in the vehicle: ddH₂O) for 48 hours. After being dehydrated by an ethanol series, paraffin was used to implant the fixed brains. Coronal sections were created using a stereotaxic atlas and a rotary microtome (RM2255, Leica, Shanghai, China) [24]. In successive 5-µm-thick coronal slices, the biggest hypothalamic areas were found, which corresponded to –3.0 mm from the bregma. The slices were examined under a light microscope (Olympus IX73, Olympus, Tokyo, Japan) after staining with thionine and hematoxylin & eosin (HE) staining.

Thionine staining was carried out as mentioned earlier [25]. Deparaffinized slices were immersed in a 4% thionine solution (Beyotime, Y267336, Shanghai, China, in the vehicle: ddH₂O) at 60 °C for 90 s. The slices were then mounted with neutral gum (Beyotime, C0173) after being dehydrated by an ethanol series.

HE staining was carried out as mentioned earlier [25]. The sections were first stained with hematoxylin (Beyotime, C0107) for 3 min, followed by immersion for 5 s in a solution of 1% hydrochloric acid alcohol. The sections were subsequently incubated with eosin (Beyotime, C0109) for 5 s. After being dehydrated by an ethanol series, the slices were then mounted with neutral gum (Beyotime, C0173).

LDH level was detected with the LDH Assay kit (Solarbio Life Sciences, BC0685, Beijing, China). Moderate hypothalamic samples were homogenized with lactic dehydrogenase assay buffer, and the LDH test kit was used to treat the supernatants after they had been separated by centrifugation. Lastly, the colorimetric test was used to detect the treated samples at 450 nm (Multiskan MS photometer type 352, Labsystems, Helsinki, Finland).

Hypothalamic samples were sliced into tissue fragments, which were then carefully dissociated into individual cell solutions using Gentle MACS™ Dissociator (Miltenyi, Cologne, NRW, Germany). The individual cell solution was then resuspended with staining buffer and treated with SYTOX green nucleic acid staining solution (Invitrogen, S34860, Carlsbad, CA, USA) at room temperature without light. The number of SYTOX-positive cells was immediately analyzed with a flow cytometer (Beckman, Brea, CA, USA).

The experimental procedure was carried out as mentioned earlier [14]. ELISA complete kits (USCN Life Science Inc., Wuhan, Hubei, China) were used to measure the levels of C-reactive protein (CRP, SEA821Ra), heat shock protein 70 (HSP70, SEA873Ra), GC (CEA540Ge), epinephrine (E, HEA858Ge), norepinephrine (NE, CEA907Ge), dopamine (DA, CEA851Ge), glucagon (Glu, CEB266Ra), angiotensin II (ANG II, CEA005Ra), interleukin-1$\beta$ (IL-1$\beta$, SEA563Ra), and IL-18 (SEA064Ra). Blood was extracted from the rats’ abdominal aorta while rats were anesthetized. After homogenizing the brain tissue, the supernatant was collected. After receiving the standard solution, serum and supernatant were incubated for 60 min at 37 °C. After adding horseradish peroxidase-labeled anti-rat CRP, HSP70, GC, E, NE, DA, Glu, ANG II, IL-1$\beta$, and IL-8 and melatonin secondary antibodies, the wells were incubated for 1.5 h at 37 °C. Following the removal of the liquid from the wells and three rounds of washing, color responses were produced using a tetramethylbenzidine color development solution. After 15 min without illumination and at room temperature, sulfuric acid (2 mmol/10 mL) was added. The plates were read using an ELISA reader (Multiskan MS photometer type 352, Labsystems, Helsinki, Finland) at 450 nm.

Protein extracts (30 µg of protein/lane) from the hypothalamus were loaded onto SDS-PAGE gels (Beyotime, P0867M), separated by electrophoresis, and transferred to PVDF membranes (Beyotime, FFP26). After incubated overnight at 4 °C with rabbit GSDME-specific monoclonal antibody (1:2000, Abcam, ab215191, Cambridge, MA, USA), rabbit GSDMD-specific monoclonal antibody (1:1500, Abcam, ab219800), rabbit GSDMD-NT-specific monoclonal antibody (1:2000, Affinity Biosciences, DF13758, Cincinnati, OH, USA), rabbit cleaved-caspse-1-specific polyclonal antibody (1:1500, Affinity Biosciences, AF4022), rabbit NLRP3-specific monoclonal antibody (1:1000, Abcam, ab263899), and mouse $\beta$-actin monoclonal antibody (1:2000, Beyotime, AF0003), the membranes were treated with horseradish peroxidase-conjugated goat anti-rabbit/mouse IgG (1:1000, Beyotime, A0208/A0216). Finally, the membranes were detected by an Odyssey gel imaging system (LI-COR, Inc., Lincoln, NE, USA).

SPSS 21.0 (IBM Corp., Chicago, IL, USA) was used for statistical analysis. All data were found to be normally distributed by the Kolmogorov-Smirnov test (p $>$ 0.1). The data are given as mean $\pm$ SEM. A post hoc least significant difference (LSD) t-test was used after the data had been analyzed using one-way or two-way ANOVA (as appropriate), to pinpoint particular group differences. p 0.05 was the significant cutoff threshold.

The behavioral changes in stressed rats were detected by open field test (Fig. 2A–E). ANOVA for the movement distance in the core region revealed significant differences [F (5, 54) = 15.03; p 0.0001]. ANOVA for the cumulative duration in the central area revealed significant differences [F (5, 54) = 45.67; p 0.0001]. ANOVA for the number of feces in the open field revealed significant differences [F (5, 54) = 13.34; p 0.0001]. Results revealed that after stress treatment, the cumulative duration and distance traveled in the core region dramatically dropped and the number of feces pellets significantly increased. Notably, these abnormal changes became more obvious with the extension of stress treatment. The present study also detected the effect of stress treatment on the body weight of rats (Fig. 2F). ANOVA for rat body weight revealed significant differences [F (5, 108) = 6.932; p 0.0001]. Unlike those of normal rats, the weight gain of rats after stress treatment was significantly reduced. Even after short-term stress treatment, the body weight of rats showed a decreasing trend.

Fig. 2.

Open-field test and body weight in Experiment 1 (n = 10). (A) The motion trajectory in the open field. (B) The movement distance in the open field. (C) The movement distance in the central area of the open field. (D) The cumulative duration in the central area of the open field. (E) The number of feces in the open field. (F) Body weight changes of stressed rats. The results are shown as the mean $\pm$ SEM, ^*p 0.05, ^**p 0.01 vs. the control group. d, day(s). RS + IS = restraint stress plus ice-water swimming.

The levels of stress-related hormones were detected by ELISA (Fig. 3). ANOVAs for serum levels of GC, E, NE, DA, Glu and ANGII were found to be significantly different from control (respectively [F (5, 54) = 33.97; p 0.0001], [F (5, 54) = 46.45; p 0.0001], [F (5, 54) = 64.12; p 0.0001], [F (5, 54) = 50.21; p 0.0001], [F (5, 54) = 43.99; p 0.0001], and [F (5, 54) = 112.2; p 0.0001]). Results showed that after short-term stress treatment, the levels of GC, E, NE, DA, Glu, and ANG II were significantly increased in rat serum. However, with the extension of stress treatment, the levels of these hormones decreased significantly.

Fig. 3.

Changes of stress-related hormone levels in Experiment 1 (n = 10). (A) The concentration of glucocorticoid (GC) in serum. (B) The concentration of epinephrine (E) in serum. (C) The concentration of norepinephrine (NE) in serum. (D) The concentration of dopamine (DA) in serum. (E) The concentration of glucagon (Glu) in serum. (F) The concentration of angiotensin II (ANG II) in serum. The results are shown as the mean $\pm$ SEM, ^*p 0.05, ^**p 0.01 vs. the control group, ^#⁢#p 0.01 vs. RS + IS group at 14 days in (A), ^#⁢#p 0.01 vs. RS + IS group at 7 days in (B), ^#⁢#p 0.01 vs. RS + IS group at 3 days in (C), ^#⁢#p 0.01 vs. RS + IS group at 7 days in (D), ^#p 0.05 vs. RS + IS group at 7 days in (E), ^#⁢#p 0.01 vs. RS + IS group at 14 days in (F). d, day(s). RS + IS = restraint stress plus ice-water swimming.

The current investigation found alterations in the concentrations of damage markers linked to stress (Fig. 4). ANOVAs for the serum level of CRP and HSP70 revealed significant differences (respectively, [F (5, 54) = 19.14; p 0.0001] and [F (5, 54) = 57.95; p 0.0001]). Results revealed that, with the extension of stress treatment, the level of CRP showed a continuous upward trend; while the level of HSP70 reacted differently, as after short-term stress treatment, the level of HSP70 was significantly increased in rat serum. However, with the extension of stress treatment, the level of HSP70 decreased significantly.

Fig. 4.

Changes in the serum levels of C-reactive protein (CRP) and heat shock protein 70 (HSP70) in Experiment 1 (n = 10). (A) The concentration of CRP in serum. (B) The concentration of HSP70 in serum. The results are shown as the mean $\pm$ SEM, ^**p 0.01 vs. the control group, ^#⁢#p 0.01 vs. RS + IS group at 14 days in (A), ^#⁢#p 0.01 vs. RS + IS group at 7 days in (B). d, day(s). RS + IS = restraint stress plus ice-water swimming.

Thionine staining detected pathological change (Fig. 5A). Pathological examination revealed that after short-term stress treatment, the cells in the hypothalamus showed no obvious histological changes; however, the number of abnormal neurons increased as the stress treatment was extended.

Fig. 5.

Pathological changes in Experiment 1 (n = 5). (A) Pathological change in the hypothalamus, as shown by Thionine staining. (a’–f’) are magnified areas of (a–f), high-power photomicrographs in the right corners of (a’–f’) are enlarged from the rectangles in (a’–f’). (B) The lactate dehydrogenase (LDH) leakage in the hypothalamus. (C) The number of SYTOX green nucleic acid staining (SYTOX)-positive cells in the hypothalamus. (D) The level of interleukin-1$\beta$ (IL-1$\beta$) in hypothalamus. (E) The level of IL-18 in the hypothalamus. (a–f) bars = 500 µm; (a’–f’) bars = 50 µm; (high-power photomicrographs) bars = 15 µm. The results are shown as the mean $\pm$ SEM, ^*p 0.05, ^**p 0.01 vs. the control group. d, day(s). RS + IS = restraint stress plus ice-water swimming.

A lactate dehydrogenase (LDH) test determines the amount of cell content leakage. SYTOX Green is a green nucleic acid dye that easily passes through injured cell membranes, but not through the cell membrane of living cells. In the present study, LDH assay (Fig. 5B) and SYTOX green nucleic acid staining (Fig. 5C, Supplementary Fig. 1A) were conducted. ANOVAs for the release level of LDH and SYTOX revealed significant differences (respectively, [F (5, 24) = 12.27; p 0.0001] and [F (5, 24) = 11.39; p 0.0001]). Results revealed that after short-term stress treatment, the level of LDH leakage and the quantity of SYTOX green acid-positive cells in the hypothalamus showed no obvious change; however, with the extension of stress treatment, a continuous upward trend was revealed for both.

When pyroptosis occurs, cells with membrane pores release interleukin-1 beta (IL-1$\beta$) and IL-18, which induces local or systemic inflammation. The levels of IL-1$\beta$ and IL-18 were detected by ELISA (Fig. 5D,E). ANOVAs for the levels of IL-1$\beta$ in the hypothalamus revealed significant differences (respectively, [F (5, 24) = 9.608; p 0.0001] and [F (5, 24) = 44.14; p 0.0001]). Results showed that after short-term stress treatment, IL-1$\beta$ and IL-18 levels showed no obvious change. However, with the extension of stress treatment, IL-1$\beta$ and IL-18 levels increased.

GSDMD and GSDME are well-known pyroptosis-related proteins. The expression status of GSDMD and GSDME in the hypothalamus was detected by Western blot (Fig. 6A). Results revealed that the expression level of GSDME in the hypothalamus was very low, but GSDMD was expressed abundantly. Necrosulfonamide (NSA) is a specific inhibitor of GSDMD that inhibits the cleavage of GSDMD, oligomerization of GSDMD-NT, and membrane perforation. NSA was used to investigate the function of the GSDMD-associated pyroptosis pathway in stress-related hypothalamic damage. ANOVA for the level of GSDMD-NT in the hypothalamus revealed significant differences [F (3, 20) = 34.58; p 0.0001]. Findings showed that stress dramatically increased the degree of GSDMD-NT in the hypothalamus and that NSA inhibited this effect (Fig. 6B,C). It is worth noting that NSA did not improve the behavioral abnormalities of stressed rats (Supplementary Fig. 2). All original WB images in Fig. 6A,B are provided in Supplementary Material 1.

Fig. 6.

Pyroptosis-related protein levels and pathological alterations in the hypothalamus in Experiment 2. (A) Representative blots showing the expression of gasdermin E (GSDME) (n = 6) and GSDMD (n = 6). (B) Representative blot showing the expression of GSDMD-N-terminal (GSDMD-NT) (n = 6). (C) Relative protein expression level of GSDMD-NT. (D) Pathological alterations in the hypothalamus, as shown by Thionine staining. (a’–d’) are magnified areas of (a–d), high-power photomicrographs in the right corners of (a’–d’) are enlarged from the rectangles in (a’–d’). (E) The LDH leakage in the hypothalamus. (F) The number of SYTOX-positive cells in the hypothalamus. (G) The level of IL-1$\beta$ in the hypothalamus. (H) The level of IL-18 in the hypothalamus. (a–d) bars = 500 µm; (a’–d’) bars = 50 µm; (high-power photomicrographs) bars = 15 µm. The results are shown as the mean $\pm$ SEM, ^**p 0.01 vs. the control group, ^#⁢#p 0.01 vs. the stress group. NSA: Necrosulfonamide, GSDMD inhibitor.

Thionine staining detected pathological changes (Fig. 6D). Pathological findings revealed that NSA alleviated stress-induced hypothalamic injury. ANOVA for the release level of LDH, the quantity of SYTOX green acid positive cells and the levels of IL-1$\beta$ and IL-18 in the hypothalamus revealed significant differences (respectively, [F (3, 20) = 23.46; p 0.0001], [F (3, 20) = 15.76; p 0.0001], [F (3, 20) = 23.65; p 0.0001] and [F (3, 20) = 81.04; p 0.0001]). Consistent with the above results, the LDH assay (Fig. 6E), SYTOX green nucleic acid staining (Fig. 6F, Supplementary Fig. 1B), and ELISA (Fig. 6G,H) revealed that NSA significantly relieved stress-induced pathological changes and inflammatory response.

Pathological alterations were observed by Thionine staining (Fig. 7A) and hematoxylin and eosin staining (Fig. 7B). In contrast to higher-dose GC (2 mg/kg and 3 mg/kg) treatment, low-dose GC (1 mg/kg) significantly alleviated hypothalamic nerve damage. ANOVAs for the release level of LDH, the number of SYTOX green acid positive cells, and the level of IL-1$\beta$ and IL-18 in the hypothalamus revealed significant differences (respectively, [F (3, 20) = 37.87; p 0.0001], [F (3, 20) = 17.81; p 0.0001], [F (3, 20) = 32.35; p 0.0001] and [F (3, 20) = 69.55; p 0.0001]). Changes in the LDH leakage (Fig. 7C), the number of SYTOX positive cells (Fig. 7D, Supplementary Fig. 1C), and the levels of IL-1$\beta$ (Fig. 7E) and IL-18 (Fig. 7F) in the hypothalamus were also significantly reversed by low-dose GC. Expression levels of the GSDMD pyroptosis pathway (cleaved-caspase-1, NLRP3, and GSDMD-NT) were detected by Western blot (Fig. 8). ANOVAs for the level of cleaved-caspase-1, level of NLRP3 and GSDMD-NT in the hypothalamus revealed significant differences [F (3, 20) = 74.52; p 0.0001], [F (3, 20) = 47.16; p 0.0001] and [F (3, 20) = 23.59; p 0.0001]). Consistent with the above results, low-dose GC significantly reversed the increased levels of GSDMD-related proteins induced by stress in the hypothalamus. All original WB images in Fig. 8A are provided in Supplementary Material 2.

Fig. 7.

Pathological changes in Experiment 3 (n = 6). (A) Pathological changes in the hypothalamus, as shown by Thionine staining. (a’–e’) are magnified areas of (a–e), high-power photomicrographs in the right corners of (a’–e’) are enlarged from the rectangles in (a’–e’). (B) Pathological changes in the hypothalamus, as shown by HE staining. (a’–e’) are magnified areas of (a–e), high-power photomicrographs in the right corners of (a’–e’) are enlarged from the rectangles in (a’–e’). (C) The LDH leakage in the hypothalamus. (D) The number of SYTOX-positive cells in the hypothalamus. (E) The level of IL-1$\beta$ in the hypothalamus. (F) The level of IL-18 in the hypothalamus. (a–e) bars = 500 µm; (a’–e’) bars = 50 µm; (high-power photomicrographs) bars = 15 µm. The results are shown as the mean $\pm$ SEM, ^**p 0.01 vs. the control group, ^#⁢#p 0.01 vs. the stress group.

Fig. 8.

The expression levels of Cleaved-cysteinyl aspartate specific proteinase-1 (Cleaved-caspase-1), nucleotide-binding oligomerization domain-like receptor protein 3 (NLRP3), and GSDMD-NT in the hypothalamus in Experiment 3 (n = 6). (A) Representative blots showing the expression of Cleaved-caspase-1, NLRP3, and GSDMD-NT. (B) Relative protein expression level of Cleaved-caspase-1. (C) Relative protein expression level of NLRP3. (D) Relative protein expression level of GSDMD-NT. The results are shown as the mean $\pm$ SEM, ^**p 0.01 vs. the control group, ^#p 0.05 vs. the stress group. GC, glucocorticoid.