Metformin Inhibits NLR Family Pyrin Domain Containing 3 (NLRP)-Relevant Neuroinflammation via an Adenosine-5′-Monophosphate-Activated Protein Kinase (AMPK)-Dependent Pathway to Alleviate Early Brain Injury After Subarachnoid Hemorrhage in Mice

Neuroinflammation plays a key role in the pathogenesis of early brain injury (EBI) after subarachnoid hemorrhage (SAH). Previous studies have shown that metformin exerts anti-inflammatory effects and promotes functional recovery in various central nervous system diseases. We designed this study to investigate the effects of metformin on EBI after SAH. Our results indicate that the use of metformin alleviates the brain edema, behavioral disorders, cell apoptosis, and neuronal injury caused by SAH. The SAH-induced NLRP3-associated inflammatory response and the activation of microglia are also suppressed by metformin. However, we found that the blockade of AMPK with compound C weakened the neuroprotective effects of metformin on EBI. Collectively, our findings indicate that metformin exerts its neuroprotective effects by inhibiting neuroinflammation in an AMPK-dependent manner, by modulating the production of NLRP3-associated proinflammatory factors and the activation of microglia.


INTRODUCTION
Subarachnoid hemorrhage (SAH) is a life-threatening subtype of stroke, characterized by high rates of mortality and poor prognoses (Laiwalla et al., 2016;Etminan et al., 2019). A previous study attributed the unfavorable prognoses of SAH patients to angiographic vasospasm (Crowley et al., 2008). However, clinical trials have shown that drugs directed against vasospasm failed to improve the poor prognoses of these patients (Macdonald et al., 2011;Macdonald et al., 2012). Recent studies have identified early brain injury (EBI), which occurs within the first 3 days of SAH, as one of the major causes of the unfavorable prognoses of SAH patients (Chen et al., 2014;Li et al., 2018). The EBI phase after SAH includes multiple pathophysiological processes. The overactivation of the inflammatory responses not only directly damages brain tissue but also contributes to the disruption of the blood-brain barrier and cell apoptosis, which further aggravate the brain injury (Fujii et al., 2013;Liu et al., 2019). A prospective observational study demonstrated that the inflammatory response is an independent predictor of unfavorable outcomes in SAH patients (Badjatia et al., 2015). Therefore, the suppression of neuroinflammation may be an effective strategy to mitigate brain injury after SAH.
The NLR family pyrin domain containing 3 (NLRP3) inflammasome is a multiprotein complex that regulates the innate immune inflammatory response (Zhou et al., 2011). The activation of the NLRP3 inflammasome causes the autocatalytic cleavage of pro-caspase 1, resulting in the processing and secretion of proinflammatory cytokines IL-1β and IL-18, which participate in the inflammatory response in multiple diseases (Mangan et al., 2018;. Recent studies have shown that the NLRP3 inflammasome contributes to the pathogenesis of EBI after SAH, and its suppression protects against brain injury (Dodd et al., 2021;Xu et al., 2021).
Adenosine-monophosphate-activated protein kinase (AMPK) is a sensor of cellular energy and modulates the status of many cellular processes (Meijer and Codogno, 2007). Recent advances in SAH research have demonstrated that the inhibition of NLRP3-inflammasome-associated neuroinflammation by activating AMPK reduces brain injury after SAH (Xu et al., 2019;Peng et al., 2020).
Metformin is widely used to treat patients with type 2 diabetes and metabolic syndrome because it safely and strongly enhances insulin sensitivity (Sanchez-Rangel and Inzucchi, 2017). A randomized controlled trial also reported that the concentrations of blood biomarkers predicting poor outcomes were significantly lower in patients with severe traumatic brain injury (TBI) treated with metformin than in untreated TBI patients (Taheri et al., 2019). Recent studies have shown that metformin exerts anti-inflammatory effects and alleviates ischemia-reperfusion injury by inhibiting the activation of the NLRP3 inflammasome (Jia et al., 2020;Zhang et al., 2020;Xian et al., 2021). In a recent study, we reported that HSP22 relieved SAH-induced EBI by salvaging the mitochondrial function in an AMPK-PGC1α-dependent manner (Fan et al., 2021). Moreover, metformin is be known to be one kind of activator of AMPK (Xie et al., 2011;Grégoire et al., 2018). To our knowledge, there have been no reports of the role of metformin in EBI after SAH. Therefore, in this study, we evaluated the effects of metformin on EBI after SAH and investigated the underlying mechanism.

Animals
Healthy adult male C57BL/6J mice (weighing 20-25 g) were obtained from the Animal Experiment Center of Southern Medical University (Guangzhou, China). The mice were housed under a 12-h light/dark cycle at a controlled temperature (22 ± 1°C) and humidity (60 ± 5%). The mice had free access to water and food. All experimental procedures involving animals were approved by the Southern Medical University Ethics Committee (Guangzhou, China) and conformed to the Guidelines of the National Institutes of Health on the Care and Use of Animals (China). The Animal Ethics Project Number is LAEC-2020-142.

Experimental Design
This study included four separate experiments, as shown in Figure 1.

Experiment 1
Western blotting was used to determine the expression patterns of phosphorylated AMPK (p-AMPK), AMPK, and NLRP3 after SAH. The mice were randomly allocated into the six groups including the sham group and five SAH subgroups (at 6, 12, 24, 48, 72 h after SAH. n = 6). The left cerebral cortices from each group were collected for western blotting analysis. Double-labeled immunofluorescence was also used for the cellular localization of p-AMPK and NLRP3 in the SAH group (at 24 h after SAH. n = 3).

Experiment 2
To determine the optimal dose of metformin for subsequent experiments, a concentration gradient of metformin was evaluated based on a previous study (Ashabi et al., 2014). Mice were randomly divided into six groups: sham, SAH, SAH + vehicle 1 (normal saline), SAH + metformin (100 mg/kg), SAH + metformin (200 mg/kg), and SAH + metformin (400 mg/kg). Metformin was administered intraperitoneally 30 min after SAH. A neurological test (modified Garcia score) (n = 8) was performed and the brain water content of the mice (n = 8) was analyzed at 24 h after SAH. After the neurological score was determined, the mice were killed and brain samples collected to assay the brain water content. The effects of different doses of metformin on NLRP3 expression in the SAH (24 h) group were evaluated with western blotting (n = 6). The SAH + vehicle 1 group was given the same volume of normal saline intraperitoneally at the same time points as the metformin-treated mice after the induction of SAH. The SAH group was only treated surgically, with no other treatment. The sham group underwent the same surgical procedure without perforation of the blood vessel, with no other treatment.

Experiment 3
The optimal dose of metformin (200 mg/kg) determined in the experiment described above was used in subsequent experiments. Nissl staining (n = 5), terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining (n = 5), immunohistochemical staining (n = 5), and western blotting (n = 6) were used to evaluate the effects of metformin on brain injury and the expression of proinflammatory factors at 24 h after SAH. The mice were randomly divided into four groups: sham, SAH, SAH + vehicle 1, and SAH + metformin (200 mg/kg).

Subarachnoid Hemorrhage Mouse Model
A mouse model of SAH was generated with endovascular perforation, as described previously (Peng et al., 2020). Briefly, C57BL/6 J mice were anesthetized with 2% isoflurane in oxygen for 3 min and anesthesia was maintained with 1.0-1.5% isoflurane in 70% N 2 O and 30% O 2 in a small-animal anesthesia system (Vet Equip, Pleasanton, CA, United States). A sharpened nylon suture was then inserted through the left external carotid artery to the common carotid artery and the internal carotid artery (ICA), ultimately perforating the intracranial bifurcation of the ICA. The sham-operated mice underwent the same procedures but the suture was withdrawn without puncture. During the operation, the body temperatures of the experimental animals were maintained at 37 ± 0.5°C with a heating pad.

Drug Administration
Metformin (Sigma, PHR1084) was dissolved in normal saline and given by intraperitoneal injection 30 min after SAH ictus. The time points and method of drug delivery were based on previous reports (Ashabi et al., 2014;Peng et al., 2020). In accordance with a previous report (Peng et al., 2020), the selective AMPK inhibitor compound C (Apexbio, B3252) was dissolved in 5% DMSO, and a working solution (5 μg in 5 μL) was injected into the intracerebroventricular cavity 30 min before the induction of SAH.

Intracerebroventricular Injection
Intracerebroventricular injections were conducted as previously described (Peng et al., 2020). Briefly, a 10 μL microsyringe (Shanghai High Pigeon Industry & Trade Co., Ltd., Shanghai, China) was inserted into the left lateral ventricle through a small cranial burr hole, which was drilled 0.3 mm posterior, 1.0 mm lateral, and 2.5 mm deep relative to the bregma. Compound C or vehicle 2 were administered slowly into the left lateral ventricle 30 min before surgery. The microsyringe was left in situ for an additional 10 min after the administration of the compound C or vehicle 2, and then withdrawn slowly. The burr hole was sealed with bone wax.

Determination of Neurological Scores and Subarachnoid Hemorrhage Grades
The neurological scores (neurobiological deficits) were determined by an independent investigator blinded to the procedural information, using the modified Garcia test (Garcia et al., 1995). The Garcia score consists of six sensorimotor tests: spontaneous activity, spontaneous movement of all limbs, forelimb outstretching, climbing, touching the trunk, and vibrissal touch. Each test was scored from 0 to 3, and the total scores ranged from 0 to 18. A higher score indicated a milder neurological deficit. The severity of SAH was evaluated blindly at 24 h after SAH, with a previously reported SAH grading system (Sugawara et al., 2008). Briefly, the brain basal cistern was divided into six parts, and each part was scored from 0 to 3 based on the amount of blood clotting present. The total score for the six parts represented the SAH grade. SAH mice with SAH grade ≤7 were excluded from this study.

Brain Water Content
At 24 h after SAH, the mice were killed and their brains quickly removed, and divided into the left hemisphere, right hemisphere, cerebellum, and brain stem. Each segment was weighed immediately to determine their wet weight (WW), and the samples were then dried at 105°C for 72 h to determine their dry weight (DW). The brain water content was calculated as [(WW − DW)/WW] × 100%.

Nissl Staining
Nissl staining was used to evaluate neuronal damage in the ipsilateral cortex, as described previously . Briefly, at 24 h after SAH, the mice were deeply anesthetized and transcardially perfused with 50 ml of phosphate-buffered saline (PBS; 0.1 M) followed by 50 ml of 4% paraformaldehyde (PFA; pH 7.4). The brain samples were removed quickly and postfixed in 4% PFA for 24 h at 4°C. After the brains were embedded in paraffin, they were cut into serial coronal sections (10 μm thick). The brain slices were deparaffinized and rehydrated. The slices were immersed in methyl violet solution (Nissl Stain Solution, G1432, Solarbio, Beijing). The slices were then dehydrated in 100% alcohol and washed with xylene. Images were obtained and analyzed under a light microscope (Leica-DM2500, Wetzlar, Germany) by a blinded investigator.

Immunohistochemical Staining (IHC)
IHC was used to verify NLRP3 and IBA1 (a microglial marker in the brain) expression in the ipsilateral cortex. The deparaffinized and rehydrated coronal brain slices (4 μm thick) were prepared as described above. The slices were incubated with 3% H 2 O 2 for 10 min at room temperature to quench any endogenous peroxidase activity, and then blocked with 5% goat serum for Frontiers in Pharmacology | www.frontiersin.org March 2022 | Volume 13 | Article 796616 20 min at room temperature. The slices were then incubated overnight at 4°C with the following primary antibodies: rabbit anti-NLRP3 (diluted 1:50; NBP1-77080SS, Novus) and mouse anti-IBA1 (diluted 1:100; GB12105, Servicebio). The next day, the brain slices were washed with PBS and incubated for 20 min at room temperature with biotinylated goat anti-rabbit IgG (diluted 1:100; ZSGB-Bio, Beijing, China) or goat anti-mouse IgG (diluted 1:100; ZSGB-Bio). The brain slices were then incubated with horseradish peroxidase (HRP)-streptavidin reagent for 10 min and stained with 3,3′-diaminobenzidine peroxidase substrate. Images were obtained with a light microscope (Leica-DM2500, Wetzlar).

Fluoro-Jade C Staining
Fluoro-Jade C (FJC) staining was used to identify degenerate neurons, as previously described (Xu et al., 2021), but with modification. Briefly, the brain slices were immersed in an alcohol solution (1% sodium hydroxide in 80% ethanol) and then in 70% ethanol, and then immersed in 0.06% potassium permanganate solution for 10 min. The slices were then incubated in a 0.0001% working solution of FJC (Millipore, Darmstadt, Germany) for 30 min. Images were obtained under a fluorescence microscope and analyzed by a blinded observer. FJC-positive cells were counted in three different fields of the ipsilateral cortex in three sections per mouse. Data are expressed as the ratio of FJC-positive cells to DAPI-positive cells.

TUNEL Staining
Cell apoptosis was detected with a TUNEL kit (Beyotime, China) at 24 h after SAH in strict accordance with the manufacturer's instructions. Briefly, the deparaffinized and rehydrated coronal brain slices were prepared as described above. After the brain slices were washed with PBS, they were incubated with TUNEL mixture for 1 h at room temperature, and then stained with DAPI. The results of TUNEL staining were observed and analyzed in the same way as FJC staining. The data are expressed as the ratio of TUNEL-positive cells to DAPIpositive cells.

Statistical Analysis
All data are presented as means ± standard deviations (SD) and all statistical analyses were performed with the SPSS 19.0 software (SPSS, IBM, Armonk, NY, United States). Student's t test was used to determine the statistical significance of differences between two groups and one-way analysis of variance (ANOVA) followed by Tukey's post hoc test was used for comparisons of multiple groups. Statistical significance was accepted at p < 0.05. Investigators were blinded to the identity of the groups during the whole experiment.

Subarachnoid Hemorrhage Model and Mortality
The total number of mice used in this study was 304, including those that died or were excluded for each group. The overall mortality rates in the sham-operated and SAH mice were 0 and 19.7%, respectively (Supplementary Figure S1B). Typical images of the brains from mice in the sham and SAH groups are shown in Figures 2A,B. Previous studies indicated that the inferior basal temporal lobe was always stained with blood and shows the most significant molecular biological changes relative to the control animal. Therefore, we mainly evaluated the basal temporal lobe of the left hemisphere adjacent to the clotted blood (Park et al., 2004;Zhang et al., 2021) ( Figure 2B). It is noteworthy that the mice treated with metformin (400 mg/kg) after SAH showed an obvious increase in death risk relative to that in the mice treated with other doses of metformin (Supplementary Figures S2A,B). Therefore, 400 mg/kg metformin was not included in subsequent experiments.

Temporal Pattern of NLRP3 and p-AMPK After Subarachnoid Hemorrhage Induction
The western blotting results indicated that the protein levels of p-AMPK and AMPK increased after SAH, peaked at 24 h, and then gradually decreased until 72 h relative to those in the sham group ( Figures 2C,D). Simultaneously, the expression of NLRP3 was similar to that of p-AMPK and AMPK ( Figures  2C,E). Double immunostaining for p-AMPK and NEUN, IBA1, or GFAP demonstrated that p-AMPK was expressed in neurons and microglia cells at 24 h after SAH. However, astrocytes did not immunostain for p-AMPK ( Figure 3A). 24 h after SAH. Astrocytes did not immunostain for NLRP3 ( Figure 3B).

Metformin Administration Improved Short-Term Neurological Functions and Attenuated Brain Edema at 24 h After Subarachnoid Hemorrhage
Besides the sham group, no significant differences were detected in the SAH grade scores among the SAH groups ( Figure 4A). The mice in the SAH group and SAH + vehicle group showed severe neurological impairment and higher brain water contents than the mice in the sham group at 24 h after SAH. The administration of both 100 mg/kg and 200 mg/kg metformin improved the neurological performance relative to that in the SAH + vehicle group. The modified Garcia test indicated that the SAH-affected mice that received metformin showed a smaller neurological deficit than the mice in the untreated SAH group, but no significant differences were observed in mice treated with different doses of metformin ( Figure 4B). The brain water content results indicated that 200 mg/kg metformin significantly mitigated brain edema relative to that in the SAH + vehicle group, whereas 100 mg/kg metformin did not ( Figure 4C). The western blotting results showed that metformin treatment significantly downregulated the SAHenhanced level of NLRP3 protein 24 h after SAH, and that 200 mg/kg metformin was superior to 100 mg/kg metformin in this regard ( Figures 4D,E). No significant differences in blood glucose were observed between the mice treated with various doses of metformin and the SAH groups (Supplementary Figure  S3). Therefore, based on all these results, 200 mg/kg metformin was selected for the subsequent experiments.

Metformin Alleviated Cell Apoptosis and Ameliorated Neuronal Injury at 24 h After Subarachnoid Hemorrhage
According to the TUNEL staining results, the sham group showed negligible TUNEL-positive cells. By contrast, the SAH group and SAH + vehicle group showed markedly increased TUNEL-positive cells. Metformin treatment reduced the proportion of apoptotic cells after SAH ( Figures 5A,C). Nissl staining was used to further assess the effects of metformin on neuronal injury after SAH. The sham group showed negligible neuronal death. The proportion of surviving neurons was significantly reduced in the SAH and SAH + vehicle groups. However, metformin treatment markedly reduced the neuronal death caused by SAH ( Figures 5B,D).

Metformin Administration Repressed the Activation of the NLRP3 Inflammasome and the Expression of the Relevant Inflammatory Cytokines at 24 h After Subarachnoid Hemorrhage
Immunohistochemical staining revealed that the expression of NLRP3 was elevated at 24 h after SAH, but this increase was inhibited by metformin ( Figure 6A). The western blotting results showed that the expressions of p-AMPK, NLRP3, cleaved caspase 1, cleaved IL-1β, and cleaved IL-18 were clearly increased in the SAH and SAH + vehicle groups relative to that in the sham group. Metformin treatment further enhanced the level of p-AMPK, but reduced the levels of NLRP3, cleaved caspase 1, cleaved IL-1β, and cleaved IL-18 relative to those in the SAH mice not treated with metformin ( Figures 6B-G).

Inhibition of Adenosine-Monophosphate-Activated Protein Kinase Weakened the Protective Effects and Anti-inflammatory Effects of Metformin at 24 h After Subarachnoid Hemorrhage
To determine whether AMPK is required for the protection against EBI offered by metformin after SAH, compound C was administered by intracerebroventricular injection. The metformin-associated alleviation of neurobehavioral deficits and the reduction in the brain water content were abolished by compound C (Figures 7A,B). FJC staining was used to further clarify the role of AMPK in metformin-ameliorated neuronal death after SAH. The inhibition of AMPK by compound C reversed the protective role of metformin against neuronal injury compared with that in the SAH + metformin group ( Figures 7C,D). The western blotting results also indicated that the effects of metformin in increasing AMPK activation and reducing the expression of NLRP3, cleaved caspase 1, cleaved IL-1β, and cleaved IL-18 were suppressed by compound C (Figures 7E-J).

The Inhibition of Adenosine-Monophosphate-Activated Protein Kinase Blocked the Inhibitory Effect of Metformin on Microglial Activation at 24 h After Subarachnoid Hemorrhage
Immunofluorescent staining showed that microglia were markedly increased in number and activated (presenting as larger cell bodies with shorter processes) in the SAH + groups compared with those in the sham group, and that the administration of metformin significantly reduced the number of IBA1-positive cells and mitigated the overactivation of microglia. However, compound C blocked these effects of metformin on the regulation of microglia, which was confirmed with immunohistochemical staining (Figures 8A-C).

DISCUSSION
In this study, we have demonstrated the beneficial effects of metformin on EBI and neuroinflammation after SAH. The important findings of our experiments are as follows. 1) The levels of p-AMPK and NLRP3 increased markedly and peaked at 24 h after SAH. Both proteins colocalized with neurons and microglia, but no fluorescent signal was detected in astrocytes.
2) The administration of metformin significantly alleviated the neurological deficit and reduced brain edema after SAH.
3) The use of metformin after SAH enhanced AMPK activation and suppressed the expression of NLRP3-relevant inflammatory mediators, thereby mitigating cell apoptosis and neuronal injury. 4) The inhibition of AMPK activation with compound C reversed the favorable effects of metformin on EBI and the inflammatory response. 5) Metformin treatment clearly suppressed microglial activation after SAH, but compound C abolished this effect. Taken together, our findings demonstrate that metformin exerts protective effects against SAH-induced EBI partly by upregulating AMPK activation, thus downregulating the NLRP3 inflammasome-associated inflammatory reaction and inhibits microglial overactivation ( Figure 8D).
An increasing number of studies support the notion that the inflammatory response is one of the key causes of EBI and is associated with severe complications after SAH (Fujii et al., 2013;Lucke-Wold et al., 2016;Wang et al., 2021). Recent studies have shown that the inhibition of NLRP3 activation significantly ameliorates SAH-induced brain injury and delays cerebral vasospasm, indicating that NLRP3-inflammasomeassociated inflammatory factors play a key role in EBI after SAH Dodd et al., 2021;Zhang et al., 2021). Consistent with a previous study Hu et al., 2021), our data show that the expression of NLRP3 is elevated and peaks at 24 h after SAH, and that NLRP3 is expressed in microglia.
Metformin has recently been reported to exert its protective effects through an anti-inflammatory mechanism by inhibiting the NLRP3 inflammasome in multiple diseases (Bullón et al., 2016;. In the present study, we found that the administration of metformin after SAH ameliorated neurobehavioral deficits, brain edema, cell apoptosis, and neuronal injury, and reduced the expression of NLRP3, cleaved caspase 1, IL-1β, and IL-18. AMPK acts as an endocellular energy sensor. Metformin is well-recognized as an activator of AMPK (Hawley et al., 2010;Hardie et al., 2012;Jin et al., 2014). In our earlier study, we demonstrated that the activation of AMPK consistently reduces brain injury by ameliorating mitochondrial dysfunction after SAH (Fan et al., 2021). Therefore, we speculated that metformin may improve brain injury after SAH by activating AMPK. In our experiments, NLRP3 and p-AMPK both localized in neurons and microglial cells, and peaked at 24 h after SAH, consistent with the results of other studies Fan et al., 2021). The similar expression trends and cellular localization of p-AMPK and NLRP3 suggested that they interact with each other. It has been reported that AMPK inhibits NLRP3 expression by activating autophagy Youssef et al., 2021). Previous studies demonstrated that AMPK activation alleviates brain injury by suppressing the activation of the endoplasmic-reticulum-stressassociated TXNIP/NLRP3 inflammasome (Li et al., 2015;Xu et al., 2019). Our results indicated that the administration of metformin enhanced the activation of AMPK, which was blocked by compound C. The protective action of metformin treatment against neurological deficit, brain edema, neuronal injury, and NLRP3-relevant inflammatory reactions after SAH were also reversed by compound C. This is consistent with recent studies that have reported that the inhibition of AMPK activation reduces the brain injury after SAH (Xu et al., 2019;Peng et al., 2020). Microglia, the resident immunocompetent and phagocytic cells of the central nervous system, produce inflammatory mediators and contribute to the neuroinflammation that occurs after SAH (Gris et al., 2019;Zheng et al., 2020). Therefore, the suppression of microglial activation limits the harmful inflammatory response and attenuates inflammation-induced EBI (Xie et al., 2017;Peng et al., 2019;Liu et al., 2020). Recent studies have provided evidence supporting the suppression of microglial activation by metformin through the activation of AMPK, both in vivo and in vitro (Pan et al., 2016;Inyang et al., 2019). Consistent with previous studies, our results indicate that metformin markedly reduces the number of IBA1-positive cells, whereas compound C abolishes this effect. Although this study demonstrates the value of metformin in attenuating EBI in a mouse model of SAH, several limitations should be noted. First, we only evaluated the anti-inflammatory characteristics of metformin, and did not examine its possible capacity to affect other regulatory mechanisms in our mouse model of SAH. Second, previous study has shown that the prolonged use of metformin after brain injury promotes angiogenesis and long-term functional recovery (Dadwal et al., 2015). Here, only the effects of a single post-SAH intraperitoneal injection of metformin were investigated, so the long-term outcomes of multiple systemic metformin treatments over different time courses should be examined in future studies.
The recent literature has hinted that metformin plays a beneficial role in improving the poor prognoses of patients with brain injury. A multicenter retrospective analysis indicated that stroke patients with diabetes who were treated with metformin showed less-severe stroke on admission and better functional outcomes at 3 months than stroke patients with diabetes who were not treated with metformin (Westphal et al., 2020). Therefore, whether metformin can improve the long-term outcomes of SAH and whether it exerts beneficial clinical effects on SAH patients must also be considered in future studies.
In conclusion, our findings support the notion that metformin exerts a neuroprotective effect by inhibiting the NLRP3-associated neuroinflammatory response and suppressing microglial activation in an AMPK-dependent manner. Therefore, metformin may provide a potential therapeutic intervention for improving EBI after SAH. Data are shown as mean ± SD. **p < 0.01 vs. sham group. ## p < 0.01 vs. SAH + Vehicle (normal saline) group. && p < 0.01 vs. SAH + Metformin + DMSO group.