Male Sprague-Dawley (SD) rats (200∼250 g) were obtained from the SPF (Beijing) Biotechnology Co., Ltd. The animals were housed with free access to food and water under a 12-h light/dark cycle. The ICH model was induced according to the previous method (Zhou et al., 2018). Briefly, rats were anesthetized (50 mg/kg pentobarbital sodium) and were positioned on a stereotaxic frame (Stoelting Co., Chicago, IL, United States). Then, 2.5 μL of 0.9% sterile saline with 0.5 U collagenase VII was injected into the right basal ganglia (Coordinate: the 1.4 mm posterior, 3.2 mm lateral to the bregma, and 5.6 mm ventral to the cortical surface). The injection lasted for over 2 min, with the needle kept in position for an additional 5 min. 2.5 μL of 0.9% sterile saline without collagenase was injected into the sham group.

All rats were randomly divided into the sham group, the ICH group, and the ICH + DEX group. The ICH and ICH + DEX group underwent the ICH procedure. The animals in the DEX-treated group received the dose of DEX with 20 μg/kg/d intraperitoneally once a day for consecutive 1, 3, or 7 days (Heng Rui Medicine Co., Ltd., China), while the sham and ICH groups were administered with the same volume of vehicle.

Two investigators who were blinded to the experimental groups evaluated the modified neurological severity score (mNSS), corner turn test, and foot-fault test. On 1st, 3rd, and 7th days post-ICH, the two investigators scored the tests independently; then, the scores were averaged. The mNSS was an integrative test to evaluate the motor, sensory, and reflex abilities. And a higher score meant a more severe injury. The corner turn test could assess the motor function. Rats were made to proceed into a corner with an angle of 30°. The direction that which individual rats exited the corner was recorded. The test was repeated ten times for each animal, and the percentage of right turns was calculated. The foot-fault test examined the placement dysfunction of forelimbs. Rats were put on an elevated grid surface and placed their paws on the wire while moving along the grid. The total number of steps and the number of foot-faults for the left forelimb the rat used to cross the grid were counted. Data were presented as the percentage of foot faults per the total number of steps.

H&E staining kit (Solarbio life science, Beijing, China) was used according to the manufacturer’s instructions. Briefly, sections were stained with hematoxylin staining solution for 10 min, Subsequently, washed with tap water for 3 min. Then, the brain sections were differentiated in 1% hydrochloric acid-ethanol for 30 s, rinsed in distilled water for 5 min, counterstained with eosin staining solution for 2 min, and dehydrated with 95% alcohol, Finally, cleared in xylene and covered with a coverslip. Then, the cellular morphology was observed under the microscope.

Nissl staining was performed to detect neuronal injury. After deparaffinization, sections were washed in PBS and incubated in Nissl’ staining solution (Beyotime Biotechnology, Shanghai, China) for about 5 min. Brain sections were then rinsed in PBS and dehydrated in gradient ethanol, cleared in xylene, and covered with a coverslip. Survival neurons were then observed under the microscope.

Fluoro-Jade C staining was applied to evaluate neurodegeneration. The sections were immersed in a solution of 1% sodium hydroxide in 80% alcohol for 5 min, 70% alcohol for 2 min, distilled water for 2 min, and followed by 0.06% potassium permanganate for 10 min with gently shaking. The sections were immersed in a solution of FJC (Millipore Corporation, United States) in acetic acid for 30 min. Then rinsed three times in distilled water and allowed to dry at 50°C for 15 min before covering and observing.

The paraffin section (5 μm-thick) of the brain was prepared, after being fixed, washed, permeabilized, and blocked, cells were incubated with rabbit anti-ZO-1 (1:200, Abcam), or rabbit anti-Iba-1 (1:100, Proteintech). After that, sections were further incubated with the fluorescein 488-conjugated anti-rabbit antibody (1:1000, Jackson Immunoresearch). Slides were imaged using a fluorescence microscope. To explore the polarization reprogramming effects of DEX, we performed immunofluorescence double labeling. Sections were incubated in mouse anti-Iba-1 (1:100, Abcam), with rabbit anti-iNOS (1:100, Abcam) or rabbit anti-CD206 (1:100, Abcam) for 1 h at 37°C. Antibodies were then used: the fluorescein 488-conjugated anti-rabbit antibody (1:1000, Jackson Immunoresearch, United States), or Cy3-conjugated anti-mouse antibody (1:1000, Jackson Immunoresearch).

Paraffin-embedded brain sections were deparaffinized and rehydrated. To assess cellular apoptosis, the sections were incubated with a proteinase K working solution (15 μg/mL in 10 mM Tris/HCl, pH 7.5) for 30 min at 37°C. After washing three times with PBS buffer, they were then incubated with 50 μL of TUNEL reaction mixture, covered with a lid, and incubated for 2 h at 37°C in the dark. Then, the slides were incubated with 100 μL stopping buffer for 10 min and then rinsed in PBS three times. DAPI was applied to detect the nuclei. Images were observed via a fluorescence microscope, and the percentage of the dUTP-positive cells was detected.

For western blot analysis, tissues adjacent to the hematoma were obtained. It was performed as described previously (Zhou et al., 2018). The following primary antibodies were used: rabbit anti-β-actin (1:6000, Abcam), or rabbit anti-Bax (1:500, Servicebio), or rabbit anti-Bcl-2 (1:1000, Proteintech), or rabbit anti-ZO-1 (1:1000, Proteintech), or rabbit anti-occludin (1:500, ABclonal), or mouse anti-TNF-α (1:1000, Proteintech), or mouse anti-IL-1β (1:1000, Abcam), or rabbit anti-IL-10 (1:700, Abcam), or rabbit anti-TGF-β (1:800, Proteintech), or rabbit anti-claudin5 (1:1000, Abcam), or rabbit anti-Iba-1 (1:1000, Cell Signaling Technology), or rabbit anti-p-IκBα (1:1000, Cell Signaling Technology) or mouse anti-IκBα (1:1000, Cell Signaling Technology). The immunopositive bands were visualized using an enhanced chemiluminescent substrate (Thermo Fisher, Waltham, MA, United States) and Bio-Rad ChemiDoc XRS digital documentation system. The amount of protein expression is presented relative to the levels of β-actin. The nuclear/cytoplasmic proteins in fresh brain tissues were isolated using the Nuclear/Cytoplasmic Fractionation Kit (Beyotime Biotechnology, Shanghai, China). Nuclear/cytoplasmic proteins were used for NF-κB p65 detection; PVDF membranes were probed with primary or rabbit anti-NF-κB p65 (1:1000, Cell Signaling Technology), rabbit anti-β-actin (1:6000, Abcam), or mouse anti-Histone H3 (1:5000, Cell Signaling Technology). The amount of cytoplasmic NF-κB p65 is presented relative to the levels of β-actin; nuclear NF-κB p65 is shown relative to the stories of Histone H3.

Evans blue (100 mg/kg) was injected into the tail vein. After 1 h, the rats were perfused with saline. The brain tissues around the hematoma were collected. After weighing, brain tissue was placed in a 50% trichloroacetic acid solution. Then, the tissues were homogenized and centrifuged (12,000 rotations/min, 20 min). The Supernatants were collected and analyzed at 620 nm using a spectrometer. EB leakage was quantified by a standard curve and was normalized by EB/brain (μg/g).

The brains of the animals were removed and weighed (wet weight). Then the brains were dried at 105°C for 24 h, and the weight of the samples was measured (dry weight). Brain water content was defined as (wet weight-dry weight)/wet weight × 100%.

All data were expressed as the mean ± SD. Repeated measures ANOVA (RM-ANOVA) was employed for behavioral tests. The remaining data were analyzed by using the student t-test and one-way ANOVA. The criterion for statistical significance was p < 0.05.

The results of the mNSS testing showed that the neurological status of the ICH group and DEX-treated group improved over time. As shown in Figure 1A, the DEX markedly ameliorated total neurological deficit scores and improved neurological functions compared to the ICH group (a decrease in the mNSS) on days 3 and 7 after ICH. ICH induction significantly increased the corner turn test compared with sham animals on days 1, 3, and 7 (Figure 1B). The corner turn test was significantly ameliorated by the administration of DEX (Figure 1B). Additionally, we conducted a foot-fault test to compare the sensorimotor behavior of the rats in the different treatment groups with that of the rats in the ICH group on days 1, 3, and 7. The rats in the ICH group showed a significantly increased number of foot faults compared with the rats in the sham group from day 1 to day 7 (Figure 1C). However, treatment with DEX reduced the number of foot faults compared with the ICH group from day 3 to day 7 (Figure 1C).

On the 3rd-day post-ICH, the morphology was observed by HE staining. The outcomes showed that the morphology of the sham group was the round and intact nucleus. Severe nuclear concentration, loose staining, and cell death could be found around the hematoma of the ICH group. The pathological degree of the DEX-therapy group in the corresponding region was prominently reduced contrasted the ICH group (Figure 1D). Subsequently, we performed Nissl’s staining to evaluate neuron viability in the perihematomal area of ICH. Compared with the sham group, the number of surviving neurons was significantly reduced after ICH (Figures 2A,B). However, after being administered with DEX, much more surviving neurons were observed (Figures 2A,B). FJC was used to evaluate the severity of neuronal degeneration (Figures 2C,D). The FJC-positive cells were observed considerably increased in the ICH group compared with the sham group. Nevertheless, the DEX treatment group markedly alleviated the number of FJC-positive cells after ICH (Figures 2C,D). To further detect neuronal cell loss and apoptosis, a TUNEL assay was conducted. ICH could induce apparent neuronal cell loss and apoptosis in the rat brain (Figures 2E,F). DEX could dramatically diminish the ratio of TUNEL⁺ staining cells (Figures 2E,F). Western blot analysis demonstrated that treatment with DEX prevented the ICH-induced Bax expression (Figures 2G,I). Moreover, DEX enhanced the anti-apoptotic Bcl-2 protein expression (Figures 2G,H).

On the 3rd-day post-ICH, Brain water content and EB extravasation were observed to assess the BBB integrity. As shown in Figure 3A, ICH remarkably enhanced the brain water content. In contrast, DEX administration dramatically reduces brain water content at 3rd after ICH (Figure 3A). Additionally, EB extravasation was elevated in the ICH group compared with the sham group (Figure 3B). Yet the rats receiving the DEX displayed a lower level of EB leakage (Figure 3B). Accumulating evidence demonstrated that the restrictive nature of the BBB is precisely due to tight junctions between adjacent endothelial cells. Hence, the tight junction proteins including ZO-1, occludin, and claudin-5 were also detected. Figures 3C–F showed that reduced ZO-1, occludin, and claudin-5 were observed in the ICH group compared with the sham group. However, the levels of the ZO-1, occludin, and claudin-5 increased in DEX group. Immunofluorescence analysis which observed the perihematomal area showed similar results of the expression of ZO-1 (Figure 3G).

Microglia/macrophages are the primary immune cells in the central nervous system and play a non-negligible role in response to ICH-induced damages. The actions of DEX on microglia/macrophage in ICH were explored on the 3rd-day post-ICH. The results of western blot analysis revealed that the expression of Iba-1 (a marker for microglia/macrophage activation) protein was significantly upregulated in the ICH compared with the sham group (Figures 4C,D). However, treatment with DEX could notably suppress the activation of microglia/macrophage (Figures 4C,D). Similarly, immunofluorescence revealed the DEX group significantly reduced the activated microglia/macrophage expression around hematomas compared with the ICH group (Figures 4A,B).

After ICH, microglia/macrophages exhibit dynamic polarization over time. We further determined the role of DEX in the modulatory effects of triggering the phenotypic conversion of microglia/macrophages, double immunofluorescent staining was examined around the hematoma of all groups. The ratio of Iba-1⁺ cells that co-expressed with M1-associated marker-iNOS was reduced by treatment of DEX (Figures 4E,F). At the same time, the percentage of Iba-1 positive cells that co-expressed with M2-associated marker-CD206 was significantly increased in the DEX-treated group (Figures 4E,G).

M1-like microglia/macrophages mainly express proinflammatory cytokines, while M2-like microglia/macrophages mainly express anti-inflammatory cytokines. An uncontrolled and prolonged M1-activated state in ICH contributes to additional neuronal damage. Compared with the sham operation group, the protein levels of M1 expression products, including TNF-α and IL-1β, and M2 expression products, including IL-10 and TGF-β, were significantly increased in the ICH group on day 3. DEX significantly decreased the production of TNF-α and IL-1β (Figures 5A–C). Simultaneously, enhanced IL-10 and TGF-β were observed in the DEX group (Figures 5D–F).

The NF-κB signaling pathway is a crucial regulator of inflammatory responses and microglia/macrophage activation. Therefore, for underlying mechanism exploration, we detected the NF-κB signaling pathway in the brain of the ICH model. The protein expression of IκBα was significantly decreased with an increase in p-IκBα in the ICH group compared with those of the sham group (Figures 6A–C). DEX caused marked dephosphorylation of IκBα (Figures 6A–C). Moreover, translocation of NF-κB p65 into the nucleus occurred after ICH (Figures 6D,E). DEX significantly blocked the translocation.