Animals

Two hundred and twenty male ICR mice (32 ± 5 g, 56 ± 5d old, Vital River, Beijing) were used in our study. The mice could forage the fodder and water ad libitum, and were housed in an animal room (20–25 ℃ temperature, 50–60% humidity) with virtual-nature 12 h light and dark cycles. All the management on the mice complies with the Rules of Laboratory Animal Care and Use, Kunming University of Science and Technology, and at the same time observed the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals.

SAH Model

Intra-arterial puncture method was adopted to establish the SAH model (Li et al. 2019). Briefly, the mouse to be operated on was anesthetized with a mixture (20 ml/kg, i.p.) of ketamine (100 mg/kg) and xylazine (10 mg/kg), and then put in a supine position. Normally, after an incision along the middle line of the neck was made, the left carotid artery (CA), the external carotid artery (ECA) and internal carotid artery (ICA) were dissected. The CA and ICA were momentarily clipped and the ECA was permanently ligated, and then a minor incision was incised on the ECA to insert a 5–0 nylon thread. After the removal of the clip, the thread went along the ICA to reach the bifurcation of middle cerebral artery (MCA) and anterior cerebral artery (ACA) at the time when contacting the obvious resistance. The thread was pushed forward approximately 2 mm and the vessel wall could be punctured. Finally, the thread was retreated, the proximal ECA was ligated and the surgical wound was sutured.

Grade of SAH

The severity of SAH was graded according to the Sugawara method (Li et al. 2019; Sugawara et al. 2008). As previously described, the basal cistern is divided into 6 areas, and each area may be scored 0 (no blood) to 3 (all arteries full of blood clot). The total score ranges from 0 – 18, and thus the severity of SAH is categorized into mild (0–7), moderate (8–12) and severe (13–18). In our study, the successful model establishment should be the severe SAH.

Intra-Cerebroventricular Injection

As was described, the anesthetized mouse was fixed to the stereotactic frame, head in a prone position. Taking the bregma the origin, the coordinate (anterior–posterior 1.0 mm, right medial–lateral 1.0 mm, dorsal–ventral depth 2.0 mm) was referred to as the injection point. The recombinant human GPNMB (rGPNMB) protein (2550-AC-050, R&D Systems) was reconstituted in phosphate buffer saline (PBS, 0.01 M) to concentration of 1 µg/10 µL, 3.3 µg/10 µL and 10 µg/10 µL, respectively. Intra-cerebroventricular injection was executed 1 h after SAH induction using a Hamilton micro-injection system at a rate of 2 µL/min; the needle stayed still for 5 min in case of the backflow and pulled back slowly. The drill was sealed with bone wax.

Intravenous Drug Administration

The selective p-AMPK inhibitor, Dorsomophin (BML-275, S7306, water soluble type, Selleck, USA) was dissolved in PBS (10 mg/kg) and injected into the tail vein under the assisted with the tail vein injection device 1 h after SAH induction.

Neurofunctional TestsModified neurological severity scores

The modified neurological severity scores (mNSS) is the most common test of neuronal deficits for mice or rats after stroke. The test is a composition of motor, balance and reflex tests. For mice, the score ranges from 0 to 14, in which 14 means the most severe neuronal deficits while 0 is intact neurofunction.

Garcia Test

Garcia test constitutes six trials, including spontaneous activity, symmetry of four limbs, forepaw outstretching, ability to climb, proprioception, and response to vibrissae. Each trial can be scored 0–3, where 0 means no function and 3 means intact function. The test score is calculated from 0 to 18.

Rotor-Rod Test

The Rotor-Rod test, either named rotarod test, is used to assess sensorimotor coordination and motor learning, and is commonly applied in the neurofunctional test in rodent models with CNS disorders. In our study, fixed rotarod was used to evaluate the mid- and long-term neuronal function. The device was set on the rotating speed at 5 revolutions per minute (RPM) and 10 RPM without acceleration (Li et al. 2020).

Morris Water Maze Test

The test was performed as we described (Li et al. 2020). All the mice were assessed at the baseline and 21 days after SAH, and continued for 6 days. On the first test day (Day 1), the platform was placed in the block (quadrant) same as the baseline. From the second test day (Day 2), the platform was removed and placed in an either clockwise or count-clockwise order to perform the test. The platform was removed on the last test day (Day 6), and the mice were free to swim for 1.5 min under the tracking system to calculate the retaining time in the target block and generate the heatmap.

Brain Water Content

Direct measure was used in our study to calculate the brain water content (BWC) (Li et al. 2019). The euthanized mouse had the brain dissected immediately without perfusion, and the cerebellum and brainstem removed. The remaining cerebrum was weighted to get the wet weight (WW). Then, it was conveyed to the electrothermostatic oven to be dried out to get the dry weight (DW). The brain water could be quantified using the percentage of the subtraction of WW and DW by the WW, which is BWC = [(WW-DW) / DW] × 100%. BWC is the quantified interpretation of brain edema.

Blood–Brain Barrier Integrity

Non-toxic Evans blue dye (1%, 10 ml/kg) was injected intravenously into the mouse (Li et al. 2019). One hour later, the deep anesthetized mouse had the transcardial perfusion with PBS (0.01 M, pH 7.40, 4 ℃). The acquired brain had the cerebellum and brainstem removed and was homogenized with 3.0 ml trichloroacetic acid (TCAA), and then centrifugated at 12,000 RPM for 30 min. The supernatant was added into the TCAA-ethanol mixture (TCAA:ethanol = 1:3) and then incubated overnight at 4 ℃. The mixture was again centrifugated at 12,000 RPM for 30 min; the resultant supernatant was examined using a spectrofluorophotometer.

Western Blot

Western blot (WB) was utilized to detect and analyze the proteins as described (Li et al. 2019, 2020). In short, the left cerebral cortex which was the same side of the surgical puncture was homogenized and centrifugated at 4 ℃ (12,000 RPM) for 30 min because this brain region has the severest SAH induced injury and the most typical pathological reactions, and the supernatant was determined using an assay kit to the equivalent concentration as per the manufacturer’s instruction (Bio-Rad Laboratories, Hercules, CA, USA). Then, the supernatant with the same protein concentration was denature in the hot water bathing at 95 ℃ for 5 min to yield the protein samples. Each time, the protein samples with same volume (10 µl) were loaded in the well of the gel to have the electrophoresis, and the gel containing the separated proteins had transfer procedure using a PVDF membrane. After protein transfer, the PVDF membrane was rinsed with Tris-buffered saline and Tween-20 (TBST) and soaked in with non-fat milk (5%) for 1 h. The primary antibodies were diluted as per the manufacturer’s instructions and added to the PVDF membrane to co-incubate overnight (12 h, 4 ℃). The PVDF membrane was then rinsed with TBST and co-incubated with secondary antibodies (2 h, 25 ℃). In completion of incubation, the membrane was rinsed and processed with ECL reagent. The final membrane was exposed on photosensitive films and the subsequent protein bands were analyzed using the ImageJ software (National Institutes of Health, MD, USA). The primary antibodies used in our study were anti-p-AMPK (1:1000, #5759, Cell Signaling Technology, USA), anti-APMK (1:1000, #5831, Cell Signaling Technology, USA), anti-p-NFκB (1:2000, ab16502, ABCAM, USA), anti-IL-1β (1:2000, ab234437, ABCAM, USA), anti-IL-6 (1:2000, ab233706, ABCAM, USA), anti-TNF-α (1:2000, ab6671, ABCAM, USA), and anti-β-actin (1:4000, ab8226, ABCAM, USA). The corresponding secondary antibodies were anti-rabbit (1:5000; ab205718, Abcam, USA) and anti-mouse (1:5000; ab205719, Abcam, USA).

Enzyme Linked Immunosorbent Assay

Enzyme linked immunosorbent assay (ELISA) was used to further confirm the expression of above inflammation related cytokines before and after SAH. The process of measurement was as per the manufacturer’s protocol and previous studies (Li et al. 2020; Zhang et al. 2019a, b). The left hemisphere cortex tissue suspension and protein samples were acquired similar to the WB method. The protein samples and the standard were mixed with antibody cocktail in pre-coated 96-well microplate, and then incubated at room temperature (25 ℃) for 1 h. After washing with buffer reagent, the development solution was added to each well, and finally the stop solution was added. Using a spectrophotometer to analyze the optical density (OD) at 450 nm for each well in the microplate, the protein concentration could be quantified by comparing the OD of target protein with standard curve. ELISA kits in the study were used: IL-1β (ab197742, Abcam, USA), IL-6 (ab222503, Abcam, USA) and TNF-α (ab208348, Abcam, USA).

Immunofluorescence Staining

The brain slides were probed using immunofluorescence (IF) staining technique (Li et al. 2019, 2020). First, the mice were deep anesthetized and had transcardial perfusion with PBS (0.01 M, pH 7.40, 4 ℃) and then paraformaldehyde (PFA) solution (4%, pH 7.40, 4 ℃). The brain was immerged in the same PFA solution for 48 h at 4 ℃. After fixation, the brain was thoroughly dehydrated in saturated sucrose solution for 72 h. Second, the brain was embedded with freezing reagent and frozen at -80 ℃ for further sectioning. Third, the hardened brain was sectioned into brain slices with the thickness of 8 µm which were mounted to the glass slides. The brain slides were washed and then incubated with Triton X-100 and bovine serum (5%) for 1 h before IF staining. The primary antibodies and the following secondary antibodies were added to the brain slices to co-incubate as described. After washing with PBS, the slides were dried and given one drop of DAPI, and then were carefully sealed with glass. Finally, the slides were taken photos under a fluorescence microscope by an independent researcher. For analyzing the expression of GBNMB, five slides from the same brain specimen shall each have four different fields (200 ×) of vision examined, focusing mainly on the nearby areas of the perforating spot on the left cortex.

Study Design

The grouping and use of mice were listed in the supplemental materials (Table S1).

Experiment 1: the trend of variation of GPNMB (n = 40).

The mice were randomly grouped into sham (n = 6), SAH 3 h (n = 6), SAH 6 h (n = 6), SAH = 12 h (n = 6), SAH 24 h (n = 6) and SAH 72 h (n = 6), 36 in total. WB was used to detect and analyze the protein expression of GPNMB before and after SAH. Four mice had IF staining to determine the expression and distribution situs of GPNMB in brain tissue.

Experiment 2: the effect of GPNMB (n = 144).

Five groups were established including sham (n = 6), SAH + vehicle (n = 6), SAH + GPNMB (1 µg/10 µL) (n = 6), SAH + GPNMB (3.3 µg/10 µL) (n = 6), and SAH + GPNMB (10 µg/10 µL) (n = 6) to study the effect of GPNMB on SAH and explore the dose–effect. SAH grading, mNSS, Garcia test, BWC and BBB integrity were assessed at 24 h and 72 h after SAH. Twenty-four mice were randomly assigned into three groups: sham (n = 8), SAH + vechicle (n = 8) and SAH + GPNMB (properly minimal dose, n = 8). The Rotarod test and Morris water maze test began to be performed before SAH (baseline) and respectively at 1 wk and 3 wk after SAH to evaluate the mid- and long-term neurofunction.

Experiment 3: the signaling pathway of the effect of GPNMB (n = 24).

Twenty-four mice were randomly grouped to sham (n = 6), SAH + vehicle (n = 6), SAH + GPNMB (n = 6) and SAH + Dorsomorphin (n = 6). WB was used to detect and analyze the protein expression of p-AMPK/AMPK, p-NFκB, IL-1β, IL-6 and TNF-α, attempting to reveal the signaling pathway by which the GPNMB affects the inflammatory response. Enzyme linked immunosorbent assay (ELISA) method was adopted to repeat the detection of aforementioned inflammatory related proteins to further confirm the signaling pathway.

Statistical Analyses

All the statistical analyses were conducted using GraphPad Prism 9.3.3 (GraphPad Software, Inc., La Jolla, USA). The test for data distribution was conducted by Kolmogorov–Smirnov and Shapiro–Wilk test. Those conforming to normal distribution were processed with one-way or two-way analysis of variances (ANOVA), the statistical results were presented as mean ± standard deviation (SD). Data with abnormal distribution were analyzed using Wilcoxon test; the results were presented as median ± standard deviation (SD). All the results were considered significant if the p value less than 0.05 (p < 0.05).