CXCL-12 Attenuates Neuroinammation via the CXCR4/PI3K/Akt Signaling Pathway in a Rat Model of SAH

Background Subarachnoid hemorrhage (SAH) is a cerebrovascular disease associated with high morbidity and mortality. CXCR4 provides a neuroprotective effect, which can alleviate brain injury and inammation induced by stroke. The purpose of this study was to evaluate the anti-inammatory effects and mechanisms of CXCR4 after SAH. Methods: SAH was induced via endovascular perforation. 185 male Sprague-Dawley rats were used. Recombinant human cysteine-X-cysteine chemokine ligand 12 (rh-CXCL-12) was administered intranasally at 1 h after SAH induction. To investigate the underlying mechanism, the inhibitors of CXCR4 and P13K, AMD3100 and LY294002, respectively, were administered intraperitoneally at 1 h before SAH. The short- and long-term neurobehavior were assessed, followed by performing western blot and immunouorescence staining. injury after SAH. The present study was the rst to investigate the neuroprotective effects of CXCL-12 and explore the potential underlying mechanisms after experimental SAH in rats. Our results demonstrated that (1) endogenous protein levels of CXCL-12 and CXCR4 were increased, and peaked at 24 h after SAH. The CXCR4 receptors were expressed on microglia at 24 h after SAH. (2) Rh-CXCL-12 improved short- and long-term neurological decits, and ameliorated brain edema at 24 h after SAH. Furthermore, rh-CXCL-12 treatment reduced the number of M1 phenotype activated microglia in the peri-hemorrhagic area in the ipsilateral cerebral cortex; (3) Administration of rh-CXCL-12 signicantly increased the expression levels of CXCR4, PI3K, and p-Akt, but decreased the expression of pro-inammatory cytokines, IL-1β, TNF-α, and IL-6, and the number of IL-1β-positive cells; (4) CXCR4 inhibitor, AMD3100, and PI3K inhibitor, LY294002, reversed the anti-neuroinammatory effects of rh-CXCL-12 and its effects on the CXCR4/PI3K/Akt signaling pathway. Taken together, our results showed that the activation of CXCR4 with rh-CXCL-12 may exert a neuroprotective effect and improve neurological functions by reducing M1 phenotype activated microglia-mediated neuroinammation after SAH, and these effects were, at least in part, via activation of the CXCR4/PI3K/Akt signaling pathway.


Introduction
Aneurysmal subarachnoid hemorrhage (SAH) is a devastating and life-threatening disease associated with high mortality and disability [1]. The annual worldwide incidence of SAH is approximately 9.1 per 100,000 people [2], and SAH resulting from intracranial aneurysm rupture accounts for 5% of all strokes [3]. Early brain injury (EBI) appears in the rst 3 days following SAH, and is the major cause of poor prognosis regarding the high mortality and delayed neurological de cits [3].
EBI begins immediately after a ruptured intracranial aneurysm, and has been proven to be the primary cause of poor outcomes after SAH. During the EBI period, ruptured intracranial aneurysm resulted in many physiological derangements, including elevated intracranial pressure, decreased cerebral blood ow, and global cerebral ischemia [4]; all of which initiate a variety of pathophysiological events, such as oxidative stress, neuroin ammation, blood-brain barrier dysfunction, and apoptosis [5,6].
Neuroin ammation induced by SAH has been considered a main devastating pathophysiological process in EBI after SAH [7].
Microglia, the resident immunecellsofthecentralnervoussystem(CNS), play an important role in regulating neuroin ammation [8]. In a healthy brain, microglia are usually in a resting status (M0). However, in the disease state, microglia adopt two different activation phenotypes: the pro-in ammatory and neurotoxic "classical" activation (M1) phenotype and the "alternative" activation (M2) phenotype with antiin ammatory and neuroprotective effects (M2) [9]. In the early stages, microglia are expressed as the M1 phenotype, which secretes proin ammatory cytokines, such as IL-1, IL-6, and TNF-α to damage brain tissue in EBI after SAH. In the late stage of SAH, the M2 phenotype plays a major role in reducing in ammation and tissue injury [10,11] Cysteine-X-cysteine chemokine ligand 12 , is an in ammatory chemokine derived from bone marrow mesenchymal stem cells, and belongs to the CXC chemokine family. CXC chemokine receptor type 4 (CXCR4) is one of seven transmembrane G-protein-coupled receptors that mediates transmembrane signaling of CXCL-12. CXCL-12 and CXCR4 are abundant and widely expressed in CNS, playing an important role in neurogenesis and contributing to the neuronal development [12]. CXCL-12 is known to be expressed in neurons, glial cells, endothelial cells, and meningeal cells [13]. Previous studies have shown that CXCL-12 is upregulated in the penumbra after stroke, reduces neurological de cits, and promotes neuroprotection and angiogenesis [14][15][16]. CXCR4 is reportedly expressed in neurons, astrocytes, microglia, and ependymal cells [17]. Recent studies have demonstrated that CXCR4 is involved in the in ammatory response and exerts neuroprotective effects after ischemic stroke [18,19].
However, there is currently no research exploring the role of CXCR4 and the underlying mechanism after SAH. The phosphatidylinositol 3-kinase (PI3K)/protein kinase B (Akt) is the downstream of CXCL-12/CXCR4 axis [20]. PI3K/Akt receives a signal from the G-protein-coupled receptor, CXCR4. P85 and P110, the two subunits of PI3K, were activated by conformational changes in space, which led to phosphorylation of Akt and activation of the pathway [21]. Recently studies have shown that the PI3K/Akt signaling pathway plays an important role in mediating cell survival, differentiation, proliferation, apoptosis, and metastasis [22]. Further evidence suggests that activation of the PI3K/Akt signaling pathway provides neuroprotective effects, which can reduce the cerebral infarct area and neuronal damage, and attenuate apoptosis in the ischemic penumbra [23]. PI3K/Akt also improves the neurobehavioral defects of SAH by reducing neuroin ammation and BBB destruction [24].
In the present study, we hypothesized that rh-CXCL-12 would attenuate neuroin ammation by inhibiting microglial activation, reducing pro-in ammatory cytokine secretion, and improving neurological impairments. Moreover, these bene cial effects were, at least in part, via up-regulation of CXCR4/PI3K/Akt signaling pathway after SAH in rats (Fig. 1).

Animals
Adult male Sprague-Dawley rats (weighting 280-330g, n=185) obtained from Guizhou Laboratory Animal Engineering Technology Center (China) were used in this project. All animals were kept in a room with controlled humidity (60± 5%), constant temperature (25± 1°C), and remained in a 12 h light and dark cycle and with ad libitum access to food and water.

SAH model
The SAH model was performed in rats using a modi ed endovascular perforation model as previously described [25]. Induction of anesthesia in rats was achieved using 4% iso urane, and was maintained using 2.5% iso urane. After intubation, the mice were placed in the supine position and connected to the rodent ventilator to breathe medical air (70%) and oxygen (30%). The heart rate, respiration, skin color, and pedal re ex assessment were assessed every ve minutes during the operation to con rm anesthesia status and prevent distress. After exposing the carotid artery and its bifurcation, a 4-0 sharp single nylon thread suture was inserted from the external carotid artery into the left internal carotid artery to the anterior and middle cerebral artery bifurcation. The nylon suture was withdrawn immediately, and iso urane was reduced to 1.5%. After the operation, the endotracheal tube was removed and the animals were placed in the heating chamber (37.5℃) to recover. Animals in the sham group underwent the same procedure, but without arterial wall puncture.

SAH Grading
The degree of SAH was assessed according to the SAH grading scale system at 24 h after SAH as previously described [26]. The basal cistern was divided into six segments that were graded from 0 to 3 according to the amount of subarachnoid blood. The total score was calculated by adding all area scores (maximum SAH grade = 18). Rats with a score of 8 or less were excluded from the current study.

Experimental design
Four separate experiments were performed as shown in Fig. 2.

Experiment 1
To determine the time course of endogenous CXCL-12 and CXCR4 protein level expression in the sham group and each group after SAH. The rats were randomly divided into six groups (n=6/group): Sham, SAH-6 h, SAH-12 h, SAH-24 h, SAH-48 h, and SAH-72 h. Western blot analysis was performed to assess the protein levels of CXCL-12 and CXCR4 in the ipsilateral (left) hemisphere cerebral cortex. Additionally, the cellular localization of CXCR4 with calcium-binding adaptor molecule 1 (Iba-1) was evaluated using double immuno uorescence staining in the Sham and SAH-24 h group (n=2/group).

Experiment 2
To evaluate the neuroprotective effects of CXCL-12 on short-term neurological outcomes after SAH, rats were randomly assigned to ve groups (n=6/group): Sham, SAH+vehicle (sterile distilled water), SAH+CXCL-12 (5 μg/kg), SAH+CXCL-12 (15 μg/kg), and SAH+CXCL-12 (45 μg/kg). CXCL-12 was administered intranasally (i.n.) at 1 h after SAH. The SAH grading score, neurobehavioral test (including modi ed Garcia test and beam balance test), and brain water content were assessed at 24 h after SAH in all groups. The best dose of CXCL-12 was selected based on the short-term neurological outcomes and brain water content results, which was also used for the following long-term outcome and mechanism experiments.
To explore the effects of CXCL-12 on microglia/macrophage activation and neutrophil in ltration at 24 h after SAH, rats were randomly assigned to three groups (n=4/group): Sham, SAH+vehicle (sterile distilled water), and SAH+CXCL-12 (optimal dose). Immuno uorescence staining of Iba-1 with CD68 and myeloperoxidase (MPO)-positive neutrophils was performed in the peri-hemorrhagic area at 24 h after SAH. Quantitative analysis of CD68-positive Iba-1 and relative uorescence density of MPO were assessed. Brain samples of these three groups were shared with experiment 4.

Experiment 3
To evaluate the effects of CXCL-12 on long-term neurobehavioral outcomes after SAH, rats were randomly assigned to three groups (n=10/group): Sham, SAH+vehicle (sterile distilled water), and SAH+CXCL-12 (optimal dose). The Rotarod test was performed on days 7, 14, and 21 after SAH. Morris water maze was performed on days 23-27 after SAH.
Drug administration CXCL-12 or vehicle was given via intranasal administration at 1 h after SAH as previously described [27].
Animals were placed in the supine position and were administered 1.5% iso urane anesthesia. A total volume of 20 μL of vehicle (sterile distilled water) or CXCL-12 (MedChem Express, NJ, USA) at three different doses (5 μg/kg, 15 μg/kg, and 45 μg/kg), with 5 μL administered every 5 minutes, alternating between the right and left nares. AMD3100 was diluted in PBS, LY294002 was diluted in 10% dimethyl sul de (DMSO), and both were administered intraperitoneally (i.p.) at 1 h before SAH.

Assessment short-term neurological performance
The short-term neurobehavioral outcomes were assessed blindly using the 18 point modi ed Garcia scoring system and the 4 point beam balance test at 24h after SAH as previously described [28]. Higher scores indicated better neurological function.

Assessment long-term neurological performance
Long-term neurobehavioral effects were assessed using the rotarod test to evaluate sensorimotor coordination and balance on days 7, 14, and 21 after SAH, and the Morris water maze was used to evaluate spatial learning capacity and memory ability on day 23-27 after SAH as previously described [29]. For the rotarod test, the animals were placed on the rotarod at the starting rate of 5 revolutions per minute (RPM) or 10 RPM, followed by gradual acceleration of 2 RPM every 5 seconds. The duration that the rats were able to stay on the accelerating rotating cylinder was recorded and used for statistical analysis. In regard to the Morris water maze test, animals were taken to the platform on the rst day of cueing test. For the spatial learning test in the following days, the animals were placed in a set of semirandom starting positions, and were tasked to nd the submerged platform within the 60-second time limit. The probe test was performed with the actual platform removed on day 27 after SAH. Swimming distance and trace, escape latency, and probe quadrant duration were recorded by the Computer Tracking System (San Diego Instruments Inc., CA, USA).

Brain water content
Brain edema was assessed by measuring brain water content using the wet-dry method as previously described [6]. The rats were euthanized at 24 h after SAH, and the brains were quickly removed and separated into four parts (right hemisphere, left hemisphere, cerebellum, and brain stem). Afterwards, each part of the brain was weighed immediately to obtain the wet weight, and then placed into an oven for 72 h at 100℃. The dried brain was weighed again. The percentage of brain water content was calculated as follows: (wet weight − dry weight) /wet weight × 100%.

Immuno uorescence staining
The rats were deeply anesthetized (5% iso urane), and were euthanized via trans-cardiac perfusion with 100-150 mL of pre-cooled PBS (4 °C) and 100 mL of 10% formalin. Whole brains were rapidly collected and xed in 10% formalin (4 °C, 24 h), followed by dehydration with 30% sucrose (4 °C, 72 h). Brain samples were embedded in OCT (Scigen Scienti c Gardena, CA, USA), and then frozen at − 80 °C. The brains were sliced into 10 μm thick coronal brain sections using a cryostat (CM3050S, Leica Next, the slices were incubated with uorescence-conjugated secondary antibodies (1:500, Jackson ImmunoResearch) for 1 h at room temperature. The slides were visualized and photographed using a uorescence microscope (DMi8, Leica Microsystems). The number of CD68-positive microglia cells were identi ed and counted in three different elds from the left basal cortex of ve random coronal sections of each rat. The positive cells were quanti ed under microscopic eld of 200x magni cation, and data were expressed as cells/ eld. To assess neuroin ammation levels, six randomly selected tubules were examined to count IL-1β-and MPO-positive cells under the microscopic eld at 400x and 200x magni cation. The uorescence intensity was quanti ed by ImageJ software (ImageJ 1.5, NIH, USA).

Western blot analysis
At 24 h after SAH, rats were deeply anesthetized (5% iso urane) and transcardially perfused with chilled PBS, followed by decapitation. The brain sections were separated into ipsilateral and contralateral hemispheres. The ipsilateral hemisphere brain tissues were snap frozen in liquid nitrogen and stored in −80 °C freezer for storage until used. Brain samples were homogenized in RIPA lysis buffer (Santa Cruz Biotechnology) with protease inhibitor for 15 min and then centrifuged at 14,000 g (4 °C, 30 min). The supernatant was collected, and protein concentration was measured by detergent compatible assay (DC Protein Assay, Bio-Rad Laboratories). Equal amounts of protein were loaded onto the 10% SDS-PAGE gel for electrophoresis and then transferred onto nitrocellulose membranes. The membranes were blocked with 5% non-fat blocking grade milk

Statistical analysis
Statistical analysis was performed using GraphPad Prism 7 (Graph Pad Software, San Diego, CA, USA). All data were presented as mean ± SD. One-way ANOVA followed by Tukey's post-hoc test was used for comparison among multiple groups. Two-way ANOVA was used to analyze the long-term neurobehavioral results. P < 0.05 was considered statistically signi cant.

Mortality and SAH grading score
Of the 185 rats used, 151 rats underwent SAH induction. Of which, 21 (14.89%) rats died within 24 h after SAH, and 10 rats were excluded from this project due to mild SAH. There was no mortality in the Sham group (Fig. s1). Subarachnoid blood clots were distributed around the circle of Willis and ventral brain stem after SAH induction, with a signi cant difference in sham group (Fig. 4 a). The average SAH grading scores among all SAH groups showed no signi cant differences (Fig. 4 b).
Expression levels of endogenous CXCL-12 and CXCR4, and colocalization of CXCR4 with microglia after SAH As shown in Fig. 3, the results of western blotting showed that the endogenous protein expression levels of CXCL-12 and CXCR4 increased in a time-dependent manner, and peaked at 24 h after SAH when compared to the sham group (P < 0.05, Fig.3 a-c). Co-immuno uorescence staining of Iba-1 with CXCL-12 showed that CXCL-12 was expressed on the microglia within cortices in the sham group and in the perihemorrhagic area at 24 h after SAH (Fig.3 d).
Intranasal administration of exogenous rh-CXCL-12 improved short-term neurobehavioral dysfunctions and attenuated brain edema at 24 h after SAH The neurobehavioral outcomes of modi ed Garcia and beam balance were signi cantly reduced at 24 h after SAH in the SAH+vehicle and SAH+rh-CXCL-12 (5 μg/kg) groups. However, administration of rh-CXCL-12 (15 μg/kg) and rh-CXCL-12 (45 μg/kg) signi cantly improved the neurological scores at 24 h after SAH (P < 0.05, Fig.4 c, d). The brain water content in the left and right hemisphere was signi cantly increased in the SAH+vehicle and SAH+rh-CXCL-12 (5 μg/kg) group, which was signi cantly reduced by the administration of rh-CXCL-12 at doses of 15μg/kg and 45 μg/kg (P < 0.05, Fig. 4 e). Brain water contents in the cerebellum and brain stem were not signi cantly different between the sham and SAH groups. Based on these results, the optimal dose of rh-CXCL-12 was 15μg/kg, which was used for the following long-term and mechanistic studies.

Rh-CXCL-12 reduced microglial activation and neutrophilic in ltration at 24 h after SAH
Iba-1 is constitutively expressed in both resting and active microglia, while CD68 is only labeled the activated microglia (M1 phenotypemicroglia). Double-immuno uorescence staining of Iba-1 with CD68 were performed to evaluate microglial activation in the ipsilateral basal cortex at 24 h after SAH. Compared to the sham group, rats in the SAH+vehicle group showed an increase in Iba-1 positive cells with activated microglial morphology of rod shape or larger body with short/thick processes. Rats treated with rh-CXCL-12 had fewer CD68-positive activated microglia at 24 h after SAH (Fig. 5 a). Quantitative analysis showed that rh-CXCL-12 administration signi cantly reduced the number of CD68-positive activated microglia (Fig. 5c). Moreover, the number of MPO-positive cells were signi cantly increased in the SAH+vehicle group (Fig. 5 a). Lastly, the rh-CXCL-12 treatment signi cantly reduced the number of MPO-positive cells (Fig. 5 d).
Rh-CXCL-12 improved long-term neurological de cits after SAH The rotarod test results revealed that the rats had signi cantly shorter falling latency at both 5 and 10 rpm in the SAH+vehicle group when compared with the sham group in the 1 st , 2 nd , and 3 rd weeks after SAH. However, administration of rh-CXCL-12 improved the rotarod performance of SAH rats compared to the SAH+vehicle group (P < 0.05; Fig. 6 a, b). The results of the water maze test showed that the escape latency and travel distance for the rats to nd the platform were remarkably increased in the SAH+vehicle group when compared to the sham group. However, a signi cantly shorter distance and time to nd the platform were observed in the SAH+rh-CXCL-12 group (P < 0.05; Fig. 6 d, e). In the probe quadrant trial, the rats in the SAH+vehicle group remained in the target quadrant for a shorter period when compared to the sham group. However, rh-CXCL-12 signi cantly increased the time spent in the target quadrant (P < 0.05; Fig. 6 c, f). There was no signi cant difference in swimming velocity among all three groups (P > 0.05, Fig. 6 g).
Rh-CXCL-12 treatment improved short-term neurological functions after SAH, and inhibition of CXCR4 and PI3K reversed such neuroprotective effects The pretreatment of CXCR4 inhibitor, AMD3100, or PI3K inhibitor, LY294002, signi cantly reversed the neurobehavioral bene ts of rh-CXCL-12 on the modi ed Garcia score and the beam balance score when compared with the corresponding control groups at 24 h after SAH (Fig. 7a, b).
Western blot results showed that the pathway-related proteins, CXCR4, PI3K, and p-Akt, and proin ammatory cytokines, IL-1β, TNF-α, and IL-6, were upregulated in the SAH+vehicle group at 24 h after SAH when compared with the sham group (Fig. 8 a-h). Rh-CXCL-12 treatment further increased the expression levels of CXCR4, PI3K, and p-Akt, but decreased the expressions of pro-in ammatory cytokines, IL-1β, IL-6, and TNF-α, compared with the SAH+vehicle group (Fig. 8 a-h). The administration of AMD3100 reversed such regulation of pathway-related proteins and the anti-neuroin ammatory effects of rh-CXCL-12 at 24 h after SAH (Fig. 8 a-h).
Administration of rh-CXCL-12 attenuated IL-1β positive cells, which were abolished by either AMD3100 or LY294002 at 24 h after SAH.

Discussion
The present study was the rst to investigate the neuroprotective effects of CXCL-12 and explore the potential underlying mechanisms after experimental SAH in rats. Our results demonstrated that (1) endogenous protein levels of CXCL-12 and CXCR4 were increased, and peaked at 24 h after SAH. The CXCR4 receptors were expressed on microglia at 24 h after SAH. (2) Rh-CXCL-12 improved short-and long-term neurological de cits, and ameliorated brain edema at 24 h after SAH. Furthermore, rh-CXCL-12 treatment reduced the number of M1 phenotype activated microglia in the peri-hemorrhagic area in the ipsilateral cerebral cortex; (3) Administration of rh-CXCL-12 signi cantly increased the expression levels of CXCR4, PI3K, and p-Akt, but decreased the expression of pro-in ammatory cytokines, IL-1β, TNF-α, and IL-6, and the number of IL-1β-positive cells; (4) CXCR4 inhibitor, AMD3100, and PI3K inhibitor, LY294002, reversed the anti-neuroin ammatory effects of rh-CXCL-12 and its effects on the CXCR4/PI3K/Akt signaling pathway. Taken together, our results showed that the activation of CXCR4 with rh-CXCL-12 may exert a neuroprotective effect and improve neurological functions by reducing M1 phenotype activated microglia-mediated neuroin ammation after SAH, and these effects were, at least in part, via activation of the CXCR4/PI3K/Akt signaling pathway.
CXCR4, a chemokine receptor in the G protein-coupled receptor gene family, is widely found in the CNS and immune cells. It induces immune cell migration and nervous system development by binding to its ligand, CXCL-12. CXCR4 is involved in regulating the in ammatory response in central nervous system diseases, such as Alzheimer's disease [30], ischemic stroke [18], and Parkinson's disease [31]. In the present study, our results showed that the expression of exogenous CXCL-12 and CXCR4 increased at the early stage of SAH and peaked at 24 h after SAH. The increased expression of CXCL-12 and CXCR4 may explain its participation in the endogenous neuroprotective mechanisms after SAH, which were insu cient in overriding the injury.
In ammatory mechanisms have been implicated in poor functional outcomes of EBI after SAH. Microglia, the resident immune cells in the brain, play a double-edged role in neuroin ammation, and are related to the functional outcome of SAH patients [32,33]. Microglia can be polarized into two different phenotypes after SAH, namely M1 and M2 [34]. M1-polarized microglia secrete numerous in ammatory cytokines, such as IL-1β, IL-6, TNF-α, and inducible nitric oxide synthase, whereas M2-polarized microglia can produce neuroprotective properties, including transforming growth factor β and IL-10 [35]. In the early stage of brain injury after SAH, microglia are mainly activated as the M1 phenotype, which aggravates in ammatory responses [11]. In the current study, we found that CXCR4 co-localized with microglia using the double immuno uorescence method, and the number of CXCR4-positive microglia was signi cantly increased at 24 after SAH, which also indicated that CXCR4 was involved in the neuroin ammation in EBI after SAH.
We then evaluated the effects of rh-CXCL-12 in the experimental SAH model. Our results showed that the intranasal administration of rh-CXCL-12, at a dose of 15 μg/kg, reduced brain water content, improved short-time neurobehavioral outcomes, which is used as the best dose of rh-CXCL-12 in further experiment.
Meanwhile, we demonstrated strong in ammatory responses and neutrophilic in ltration, as evidenced by increased M1 microglia/macrophage polarization (CD68-positive microglia) and MPO-positive cells, which was consistent with previous investigations [28]. Intranasal administration of rh-CXCL-12 reduced the number of M1 phenotype activated microglia and MPO-positive cells. In addition, neuronal apoptosis and the in ammatory response induced hippocampal neuronal degeneration, leading to the decline of memory and space learning ability after SAH [36]. In the current study, we found that rh-CXCL-12 notably improved long-term cognitive and memory impairment, which indicated that rh-CXCL-12 exerted a neuroprotective role in delayed brain injuries induced by SAH.
We further investigated the underlying molecular mechanism of CXCL-12-induced anti-in ammatory effects after SAH. The intracellular PI3K/Akt signaling pathway plays a key role in regulating the cell cycle, including cell proliferation, cellular quiescence, cancer, and longevity. Activation of PI3K results in the phosphorylation of Akt, which further participates in various molecular cascades. Phosphorylation of Akt maintains mitochondrial integrity and promotes cellular survival by resisting in ammation and oxidative stress-induced damage [28]. Many studies have found that the PI3K/Akt pathway is closely related to cerebral ischemic and hemorrhagic injury [36,37]. Sugawara et al. [38] found that intraperitoneal injection of simvastatin reduced nerve dysfunction and cerebral vasospasm in SAH rats, and intravenous injection of wortmannin (PI3K/Akt inhibitor) abolished the neuroprotective effects of simvastatin, which indicated that the PI3K/Akt signaling pathway is involved in the neuroprotective effect of cerebral vasospasm after SAH. Zhang HB et al. [39] demonstrated that propofol attenuates SAHinduced EBI by inhibiting in ammation and oxidative stress, which was reversed by LY294002 (PI3K/Akt inhibitor) administration. Furthermore, broblast growth factor-2 alleviated neurological impairments, brain edema, and neuronal apoptosis following SAH. PD173074 (PI3K/Akt inhibitor) abolished the antiapoptotic effects of broblast growth factor-2 via suppression of the expression of PI3k [27]. These results indicate that PI3K/Akt signaling participates in anti-apoptosis and anti-neuroin ammation after SAH. Recent studies demonstrated that CXCL-12/CXCR4 could promote angiogenesis and regulate cell apoptosis by activating the PI3K/Akt signaling pathway, which indicated that the PI3K/Akt signaling pathway was a downstream molecule of the CXCL-12/CXCR4 axis [40,41]. In this study, we found that rh-CXCL-12 treatment signi cantly improved the modi ed Garcia and beam balance scores, and increased the expression levels of CXCR4, PI3K, and p-Akt, but decreased the expression of pro-in ammatory cytokines, IL-1β, TNF-α, and IL-6, as well as IL-1β-positive cells. Furthermore, CXCR4 inhibitor, AMD3100, and PI3K inhibitor, LY294002, reversed the neuroprotective effects of rh-CXCL-12 by decreasing the levels of IL-1β, TNF-α, and IL-6. Taken together, rh-CXCL-12 attenuated neuroin ammation, suppressed M1microglial polarization, and alleviated neutrophilic in ltration, which functioned, at least in part, by activating the CXCR4/PI3K/Akt signaling pathway after SAH.
This study has several limitations. First, this study focused primarily on the role of neuroin ammation in the EBI phase, and assessed the potential mechanisms of in ammatory responses at 24 h after SAH. However, neuroin ammation in delayed brain injury after SAH should also be investigated in future studies. Second, in addition to neuroin ammation, the pathogenesis of SAH includes neuronal apoptosis, oxidative stress, destruction of the blood-brain barrier, etc. However, the present experiment only focused on neuroin ammation. Therefore, the observation and study of other mechanisms should be elucidated in future experiments. Third, CXCR4 has other downstream signaling pathways, such as MEK/ERK [42].
Therefore, more experiments are necessary to investigate the possible mechanisms of these signaling pathways and the neuroprotective effects of CXCL-12.

Conclusions
Our results demonstrated that the activation of CXCR4 with CXCL-12 improved short-and long-term neurological de cits, and attenuated neuroin ammation in EBI after SAH in rats. The protective effects of CXCL-12 were at least in part through activation of the CXCR4/PI3K/Akt signaling pathway. Therefore, early administration of CXCL-12 may provide a therapeutic strategy against brain injury after SAH.