Neurogenesis and Proliferation of Neural Stem/Progenitor Cells Conferred by Artesunate via FOXO3a/p27Kip1 Axis in Mouse Stroke Model

Promoting neurogenesis and proliferation of endogenous neural stem/progenitor cells (NSPCs) is considered a promising strategy for neurorehabilitation after stroke. Our previous study revealed that a moderate dose of artesunate (ART, 150 mg/kg) could enhance functional recovery in middle cerebral artery occlusion (MCAO) mice. This study aimed to investigate the effects of ART treatment on neurogenesis and proliferation of NSPCs using a rodent MCAO model. MRI results indicated that the ischemic brain volume of MCAO mice was reduced by ART treatment. The results of diffusion tensor imaging, electron microscopic, and immunofluorescence of Tuj-1 also revealed that ischemia-induced white matter lesion was alleviated by ART treatment. After ischemia/reperfusion, the proportion of Brdu + endogenous NSPCs in the ipsilateral subventricular zone and peri-infarct cortex was increased by ART treatment. Furthermore, the neuro-restorative effects of ART were abolished by the overexpression of FOXO3a. These findings suggested that ART could rescue ischemia/reperfusion damage and alleviate white matter injury, subsequently contributing to post-stroke functional recovery by promoting neurogenesis and proliferation of endogenous NSPCs via the FOXO3a/p27Kip1 pathway.


Introduction
As recently reported by the American Heart Association (AHA), the mean global lifetime risk of stroke has increased by 8.9% from 1999 to 2016 [1]. Post-stroke neurologic deficit causes a massive social and family burden, while neuro-functional recovery has been intensively investigated [2]. Neural stem/progenitor cells (NSPCs) can be self-renewal and differentiate into mature neurons, oligodendrocytes, and astrocytes [3]. Previous studies have indicated that activating the endogenous NSPCs could be a novel cell therapy approach for hypoxic-ischemic brain injuries [4]. However, the neurogenesis and differentiation of endogenous NSPCs are remarkably disrupted in an impaired neurovascular microenvironment after hypoxicischemic brain injuries [5]. Artesunate (ART), a derivative of artemisinin, exhibits high efficacy and low toxicity in malarial treatment [6]. ART was approved for first-line medical use of malaria by the US Food and Drug Administration in 2020 for its low mortality and high toleration than quinine in several trials [7,8]. The ability to penetrate the blood-brain barrier is commonly used in patients with cerebral malaria [9]. ART's potential tumor-suppressive roles have been widely studied [10], as well as its effects on organ injury [11], immune responses [12], and neurotransmission regulation [13]. Our previous study has suggested that ART restores the damaged blood-brain barrier by regulating the S1pR/ PI3K signaling pathway in a mouse model of subarachnoid hemorrhage [14]. Moreover, our recent findings have revealed that the proliferation of NSPCs and neurogenesis are promoted by ART-modulated PI3K/Akt/FOXO-3a/ p27 kip1 signaling in an in vitro oxygen/glucose deprivation model [15]. However, ART treatment's neuroprotective roles on endogenous NSPCs in the subventricular zone (SVZ) and white matter injury have not been investigated using an animal stroke model. FOXO transcriptional factors are the targets of the PI3K/Akt signaling pathway [16]. Upon dephosphorylation, FOXO3a is translocated into the nucleus to induce the expression of the cyclin-dependent kinase inhibitor 1B (p27kip1) [17]. Subsequently, the formation of a cyclinE-CDK2 complex induced by p27kip1 could trigger cell cycle arrest at G0/G1 phase, suppressing neurogenesis and proliferation of NSPCs [18]. Inhibition of FOXO3a/p27 Kip1 signaling could break the cell cycle arrest in endogenous neural stem/progenitor cells induced by ischemia/reperfusion. Moreover, accumulated evidence showed the resting endogenous NSPCs in SVZ were activated upon acute ischemic attack, migrated toward the lesioned area, and directed primarily into reactive astrocytes, which contributed to scar formation [19]. The glial scar formation in the penumbral region around the core could block the neurorepair from ischemia insult. Doublecortin (DCX) is preferentially expressed in neuronal precursor cells and immature neurons in embryonic and adult cortical structures [20]. DCX + cells are restricted into the neuronal lineage to repair the cortical thalamic tract and cannot differentiate into glial cells [21]. Therefore, DCX/Brdu co-staining was adopted in our present experiment to show whether the ART-induced NSPCs differentiate into newborn neurons instead of glial cells in the peri-infarct cortex.
Based on these findings, we hypothesized that ART could rescue the penumbra damage, alleviate white matter injury, and reverse neurologic impairment after ischemic stroke. The present study aimed to investigate ART's potential neuroprotective functions and the underlying mechanisms using an MCAO mouse model.

Establishment of Animal Model
A total of 348 C57BL/6 J male mice (12 to 16-week-old; 25-28 g; Jackson Laboratory, stock no. 000664) were purchased from the Animal Center of the Third Military Medical University. Mice were housed in a pathogen-free environment and had free access to food and water with the light/ dark cycle of 12:12 h.
As previously described, the MCAO model was used to generate focal unilateral cerebral ischemia [22]. Briefly, C57BL/6 mice were anesthetized using a 2% isoflurane/air mixture (1-2 L/min). A vertical midline incision exposed the submandibular glands on the neck. The carotid sheath was removed, and the common carotid artery was separated to avoid accidental injury to the vagus nerve. A 2.0-cm silicone-coated nylon suture with a diameter of 0.23 mm (Sunbio Co., Ltd., Beijing, China) was gently inserted from the external carotid artery stump internal carotid artery, which terminated at the beginning of the middle cerebral artery. The cerebral blood flow was blocked for 90 min. The same procedures were performed on sham group mice apart from occlusion. The temperature of experimental animals was monitored during the surgery and the recovery from anesthesia. All surgeries and tests were carried out randomly by blinded researchers.

Experiment 1
The experimental mice were randomly sorted into five groups to evaluate the recovery of motor function in MCAO animals treated with different ART concentrations (n = 10 in each group): Sham, MCAO + Veh (saline); MCAO + 50 mg/ kg ART; MCAO + 150 mg/kg ART; and MCAO + 250 mg/ kg ART. The ART (Holley-Wulingshan pharmaceuticals Ltd., Chongqing, China) was given intraperitoneally once per day, and the first dose was given immediately after reperfusion. Furthermore, each group's recovery of forelimb motor function was assessed using cylinder task, grid-walking task pasta handing, and single-pellet retrieval task. Animals were tested once on these tasks one week before surgery to establish baseline performance levels and were then tested at 3 and 7 days post-MCAO. Behaviors were scored by observers blinded to the grouping (Fig S1A of Supplemental Material). Detailed descriptions were presented in supplementary materials.

Experiment 2
To further examine the efficacy of ART and investigate the underlying mechanisms of its neuroprotective roles, the established MCAO model mice were randomly grouped (n = 6 in each group of different tests): Sham, MCAO + Veh (saline; Ad-Flag,), MCAO + ART (150 mg/kg; Ad-Flag) and MCAO + ART + FOXO3aOE (150 mg/kg/day; Ad-TM-FOXO3a). T2-weight MRI was used to measure the cerebral infarcted volume. Diffusion tensor imaging (DTI), transmission electron microscopy, and immunofluorescence assay were performed to examine white matter injury. Immunofluorescence staining and western blotting (WB) were carried out to assess the proliferation of NSPCs in the SVZ and perilesional cortex, which revealed the proliferation and migration of NSPCs after ischemia/reperfusion. FOXO3a/ p27kip1-associated molecules' expression levels were determined three days post-MCAO (Fig S1B of Supplemental Material).
To further investigate ART's effects on the stroke model, the three experimental groups, sham, MCAO + Veh, and MCAO + ART, were assessed by comparing neuro-function, infracted volume, and white matter injury. Additionally, MCAO + ART and MCAO + ART + FOXO3a OE groups were also examined to elucidate the underlying mechanisms of ART-induced neuroprotective roles.

Magnetic Resonance Imaging and Diffusion Tensor Imaging
Mice were anesthetized using a 2% isoflurane/air mixture before magnetic resonance imaging (MRI). Serial and T2-weighed imaging was carried out on 16 coronal slices (thickness = 1.0 mm) using a 7.0-T Varian MR scanner (field of view = 35 × 25 mm; Bruker Analytik GmbH, Karlsruhe, Germany) at 72 h after MCAO.
Diffusion tensor imaging (DTI) assay enabled the visualization and characterization of white matter fibers in three dimensions and was performed as previously described [23]. Diffusion Toolkit software (Harvard Medical School, Boston, USA) was used to track white matter fiber bundles and calculate apparent diffusion coefficient (ADC) and the fractional anisotropy (FA) values in the ipsilateral internal capsule. The FA value (Ipsil/contral) was calculated as (FA value at ipsilateral IC)/ (FA value at contralateral internal capsule) × 100%.

Transmission Electron Microscopy
At 72 h after MCAO, mice were anesthetized and perfused with pre-cooled 0.01 M PBS and 4% formaldehyde. After the perfusion, brain tissues were obtained and immersed in glutaraldehyde. The per-infracted cerebral cortex was immediately removed from the solution. Subsequently, the white matter around the cerebral infarction was homogenized into small pieces (~ 1 mm 3 ) and then stored in 2% formaldehyde and 2% at glutaraldehyde at 4 °C overnight. After dehydration, samples were impregnated with epoxy resin, sectioned, and then double-stained using lead citrate and uranyl acetate. Images were captured with an H-7100 transmission electron microscope (Hitachi, Tokyo, Japan) as previously described [25].

Bromodeoxyuridine Injection and Incorporation Assay
One hundred milligram/kilogram of bromodeoxyuridine (BrdU) was administered intraperitoneally every 24 h, and the first dose was given immediately after MCAO. The coronal sections (thickness = 25 μm) were stained. The confocal images visualized the staining of BrdU and DCX doublepositive cells in the peri-infarct and SVZ regions. Cell counting was performed on four slices for each brain to calculate the relative percentage of stained cells in the two areas.

Adenovirus Transfection
An adenoviral vector expressing constitutively active FOXO3a(AD-TM-FOXO3a) was adopted; thus, TM-FOXO3a could not be phosphorylated by p-Akt. Adenovirus Ad-TM (triple mutant)-FOXO3a (FOXO3a OE ) and negative control Ad-Flag were purchased from Shanghai Heyuan Biological Co., Ltd. In the FOXO3a overexpression vector, three conserved Akt phosphorylation sites (Thr-32, Ser-253, and Ser-315) of FOXO3a were replaced by alanine residues to prevent phosphorylating by p-Akt. Ad-TM-FOXO3a and Ad-Flag (2 μL, 10 7 pfu) were injected into the lateral ventricles (0.33 μl/min) using a specific coordinate (0.2 mm posterior to the bregma, 2 mm ventral) the skull, and 1 mm lateral to the sagittal line) with stereotaxic frame 48 h before MCAO surgery.

Statistical Analysis
All data were presented as the mean ± standard error (SEM) and passed a normality test. Comparisons between two groups were analyzed using 2-tailed Student t-tests. Behavioral data collected at repeating time points were analyzed using two-way repeated-measures ANOVA, followed by the Bonferroni post hoc test. Using one-way ANOVA followed by Tukey's post hoc test, other data were analyzed. The statistical analyses were conducted using SPSS v19.0 (SPSS Inc., Chicago, IL). P < 0.05 was considered statistically significant.

ART Alleviated the Motor Dysfunction, Reduced Cerebral Infarct Volume, and White Matter Lesion After MCAO
The results of cylinder task, grid-walking task, and pasta handing, single-pellet retrieval task (Fig S3 of Supplemental Material) indicated that only 150 mg/kg ART treatment remarkably promoted functional recovery at day 3 and 7 after ischemia/reperfusion. Those findings suggested that 150 mg/kg ART could improve the final fine motor function outcome in the sub-acute phase of stroke. Moreover, animals' survival rate and body weight in the MCAO + ART group were not reduced compared to the MCAO + Veh group (p > 0.05, Fig S2 of Supplemental Material).
The T2-weighted images revealed that the ischemic size of MCAO mice was reduced by ART treatment three days post-reperfusion (p < 0.05, Fig. 1A, B). The ADC and FA values were calculated using three diagonalized diffusion tensor elements in the ipsilateral internal capsule (IC). Significant reduction and fragmentation of nerve fiber bundles were observed in the ipsilateral hemisphere three days following MCAO (Fig. 1C). Compared with the sham group, ADC value was remarkedly elevated, and FA value was notably decreased in infarcted IC at day 3 (p < 0.05), suggesting nerve fiber damage in ipsilateral IC. Moreover, 150 mg/kg of ART prevented white matter lesions induced by MCAO, as the reduction of ADC and FA values was reversed in the MCAO + ART group compared to the MCAO + Veh group (p < 0.05; Fig. 1C-E).
The axon and myelin sheath injury at the infract IC at day 3 after MCAO were analyzed by electron microscopy, which indicated that treatment with 150 mg/kg ART alleviated axon swelling and myelin injury in infracted white matter fiber (Fig. 1F). Besides, the mean axonal number was significantly increased by ART treatment (p < 0.05; Fig. 1G), whereas mean G-ratio (ratio of axon diameter to total outer diameter) (p < 0.05; Fig. 1H) and axon diameter (p < 0.05; Fig. 1I) were reduced compared with MCAO + Veh group. Furthermore, immunofluorescence staining of Tuj-1 revealed many axons and fractured dendrites, as well as an infarcted internal capsule at day 3 post-MCAO (p < 0.05; Fig. 1J). The number of Tuj-1 + cells and the fluorescence intensity of Tuj-1 were significantly increased in the MCAO + ART group compared to the control (p < 0.05; Fig. 1K, L).

ART Promoted Post-MCAO Neurogenesis Through FOXO3a/p27 Kip1 Pathway
Normalized DCX intensity was significantly increased in ipsilateral SVZ by ART treatment (p < 0.05; Fig. 2A, B). Furthermore, the numbers of DCX + and Brdu + cells were measured to evaluate the proliferation of NSPCs after MCAO in the ipsilateral SVZ ( Fig. 2A, C). The result indicated that 150 mg/kg of ART significantly promoted neurogenesis in ipsilateral SVZ after MCAO (p < 0.05). Furthermore, the BrdU incorporation assay was carried out to evaluate the proliferation of DCX + cells in the peri-lesioned cortex. Compared with the MCAO + Veh group, ART treatment notably increased the number of DCX + and Brdu + cells and elevated DCX intensity in the perilesional cortex on day 3 after MCAO (p < 0.05; Fig. 2D-F). Furthermore, Nestin's expression levels, a specific marker for NSPCs, were significantly increased by ART treatment in the peri-infarct cortex (p < 0.05; Fig. 3A, F).
The result of WB in the peri-infarct cortex suggested that ART treatment decreased the expression of FOXO3a (p < 0.05), which was reversed by the treatment with Ad-TM-FOXO3a (FOXO3a OE ; p < 0.05, Fig. 3A, B). Additionally, the expression levels of p27 Kip1 were investigated, and the results indicated that ART treatment significantly reduced the expression of p27 Kip1 (p < 0.05), and these effects were abolished by FOXO3a OE (p < 0.05, Fig. 3C). Our WB data revealed that the expression levels of Cyclin E and CDK2 were remarkably upregulated by ART treatment (p < 0.05), which was abolished by FOXO3a OE (p < 0.05, Fig. 3D, E). Moreover, Nestin's ART-induced expression in the peri-infarct cortex was also reversed by the treatment with FOXO3a OE (p < 0.05; Fig. 3F). Therefore, these WB data revealed that ART could promote post-MCAO neurogenesis in mice by regulating the FOXO3a/p27 kip1 /Cyclin E/CDK2 pathway.

The Roles of FOXO3a/p27 Kip1 Signaling on ART-Induced Neuroprotection in MCAO Mice
The expression of FOXO3a in the mice's brain of the Ad-TM-FOXO3a group was significantly higher than the Ad-Flag group (p < 0.05; Fig S4 of Supplemental Material). The cerebral infarct volume was evaluated by T2-weighed imaging (Fig. 4A) and TTC staining (Fig. 4B) three days after MCAO, and overexpressed FOXO3a reversed ART-induced reduction of cerebral infarct volume (MCAO + ART + FOXO3a OE vs. MCAO + ART group, p < 0.05; Fig. 4C, D). Exceptional motor function and sensitivity to injury of MCAO were also evaluated using the grid-walking and single-pellet retrieval tasks. The results indicated that overexpression of FOXO3a notably abolished the protective effects of ART treatment at 3 and 7 days post-MCAO (p < 0.05; Fig. 4E, F).
Also, white matter lesions in the infarcted brain were examined by transmission electron microscopy (Fig. 5A). The results indicated that overexpressed FOXO3a significantly abolished the protective effects of ART on axon swelling and myelin injury in white matter fiber, and ART-induced impact on the mean axonal number, G-ratio, and axon diameter remarkably were also reversed (MCAO + ART + FOXO3a OE vs. MCAO + ART group, p < 0.05; Fig. 5B-D). Post-MCAO neurogenesis was also evaluated by immunofluorescence staining of DCX and BrdU (Fig. 5E). The ART-induced increase of normalized DCX intensity and the numbers of DCX + and Brdu + cells in the peri-lesioned cortex were abolished by overexpressed FOXO3a at day 3 post-MCAO (p < 0.05; Fig. 5F). Moreover, increased BrdU intensity in ipsilateral SVZ caused by ART was also abolished by the overexpression of FOXO3a (p < 0.05; Fig S5 of Supplemental Material).

Discussion
Post-stroke functional recovery remains challenging, posing a substantial threat to public health [26]. Therefore, it is essential to develop therapeutic strategies for patients with ischemic stroke. ART, a stable derivative of artemisinin with high efficacy and low toxicity, exerts anti-malarial functions and can penetrate the blood-brain barrier [27]. Apart from its anti-neuroinflammatory and antioxidant roles, ART can also protect the blood-brain barrier and promote neurogenesis; therefore, it is a novel candidate for treating neurological disorders, according to our previous report [28].
In Experiment I, three concentrations of ART (50, 150, and 250 mg/kg) were used to elucidate the effects of different doses on motor function recovery in MCAO mice. The recovery of each group's forelimb motor function was assessed using cylinder, grid-walking, and pasta-handling, single-pellet retrieval tasks. We found that ART at 150 mg/ kg significantly promoted the delicate finger functional Fig. 4 FOXO3a was involved in ART-modulated cerebral infarct volume change and motor dysfunction. (A, B) The cerebral infarct volume was visualized by T2-weighed imaging and TTC staining three days after MCAO. (C, D) Quantification of cerebral infarct volume. Overexpression of FOXO3a reversed ART-induced reduction of cerebral infarct volume. n = 6 in each group. (E, F) The gridwalking task and single-pellet retrieval task were performed to examine fine motor function. n = 6 in each group. #p < 0.05 vs. MCAO + ART group. ART, artesunate; FOXO3a OE , overexpression model of FOXO3a generated using mutant type Ad-TM-FOXO3a recombinant adenovirus vectors recovery at the sub-phase of stroke (Fig S3 of Supplemental Material). The cerebral infarcted volume and white matter injury were examined to assess ART's efficacy in the MCAO model. The results from T2-weight MRI indicated that the ischemic brain volume of MCAO mice was reduced by ART treatment three days after reperfusion (Fig. 1). The white matters, including the axon and myelin of cortical projection neurons, are vulnerable to ischemia/ reperfusion lesion, which results in contralateral motor function deficiency after MCAO. Our results from DTI, electron microscopy, and immunofluorescence staining of Tuj-1 suggested that 150 mg/kg of ART also alleviated post-MCAO white matter lesion in mice (Fig. 1). Furthermore, the proliferation and neurogenesis of NSPCs in the peri-infract region were significantly promoted by ART treatment at day 3 post-MCAO (Fig. 2). The bystander effects [29] of NSCPs may serve essential roles in reducing infarct volume in the early stroke stage, revealed by TTC staining and MRI (Fig. 4). The primary functions of NSCPs include protecting the blood-brain barrier [30], promoting the secretion of neurotrophic factors such as BDNF, NTFs, and VEGF [31], inhibiting local neuroinflammation [32], and enhancing angiogenesis [33]. Therefore, activation of NSPCs by ART treatment could reduce the infarct volume at the early stage of stroke. Moreover, previous studies have suggested that DCX is a microtubule-associated protein expressed by immature neurons or neuronal-restricted precursors that could not differentiate into glial cells [21]. In the present study, whether ART could promote post-MCAO neurogenesis was investigated. The numbers of DCX and BrdU double-stained cells in the ipsilateral SVZ and peri-lesioned cortex were compared three days after MCAO. The results indicated that ART could activate endogenous NSPCs to proliferate and differentiate rapidly in both the ipsilateral subventricular zone and perilesional cortex, subsequently reconstructing the neural network of the ischemic area (Fig. 2). The previous study has also revealed that dead neurons in the infarct cortex are replaced by regionally-induced NSPCs from cortical and subpial regions and leptomeninges (pia mater), rather than migration of NSPCs from the SVZ [34].
Furthermore, this review suggested that activation of NSPCs could induce neurogenesis in the peri-ischemic areas but not in the ischemic core. Our data revealed that the proliferation and neurogenesis of endogenous NSPCs in SVZ were promoted by ART treatment and activated NSPCs relocated from SVZ to the peri-ischemic areas in MCAO mice ( Fig. 2A). Furthermore, ART-induced neurogenesis in the peri-ischemic regions was confirmed by the co-staining of DCX and Nestin (Figs. 2D and 5E). Although the proliferating immature neurons in the peri-infract area could be partially from SVZ-derived NSPCs, our data still suggested that ART could promote the proliferation and neurogenesis of SVZ-derived or in situ NSPCs after stroke.
FOXO3a, a key transcription factor, is phosphorylated by Akt. FOXO3a is dephosphorylated and translocated into the nucleus to regulate multiple gene transcription [35]. Cyclin-dependent kinase inhibitor 1B (p27 Kip1 ), which FOXO3a holds, belongs to the Cip/Kip family of cyclindependent kinase (CDK) inhibitor and binds to cyclin E, consequently downregulating cyclin E and CDK2 at G1/ G0 phase [36]. When NSPCs are stimulated by hypoxia, FOXO3a is dephosphorylated and translocated into the nucleus to induce the p27 Kip1 expression; thus, the cell cycle of NSPCs will be arrested at G0/G1 phase [37]. The expression levels of FOXO3a/p27 kip1 -associated molecules were determined using WB in MCAO mice to investigate the underlying mechanisms of ART-induced proliferation in NSPCs. The activity of protein kinase complex CyclinE-CDK2 could be inhibited by cyclin-dependent kinase inhibitor 1B (p27 Kip1 ), a KIP family member, and regulated by FOXO3a. Our results revealed that ART treatment caused inhibition of FOXO3a and downregulation of p27 Kip1 , which subsequently promoted endogenous NSCPs proliferation in the SVZ and peri-infarct zone of MCAO mice (Figs. 2 and  3).
To identify FOXO3a/p27Kip1 pathway involvement on ART-modulated neuroprotective roles in MCAO mice, T2-weight MRI, TTC assay, transmission electron microscopy, and immunofluorescence assay were performed to investigate whether overexpressed FOXO3a could reverse the neuroprotective effects of ART. Improved performance in behavior tests and reduced cerebral infarct volume caused by ART treatment were significantly abolished by FOXO3a OE (Fig. 4). Alleviated white matter lesion induced by ART treatment was also abrogated by FOXO3a OE (Fig. 5A-D). Moreover, ART-induced normalization of DCX intensity was abolished by FOXO3a OE treatment (Fig. 5E, F), suggesting that ART promoted post-MCAO neurogenesis in mice through FOXO3a/ p27 Kip1 signaling.
However, there were still some limitations in the present study. For example, ART treatment's effects and safety should be further investigated on the multiple animal stroke models, especially in nonhuman primates. Moreover, a comparison of the activating endogenous NSPCs and transplanting exogenous NSPCs should be performed in further study. We are focusing on the repair effects of the endogenous NSPCs neurogenesis and proliferation in the current study.
Neurogenesis and differentiation of endogenous NSPCs in the adult brain are considered a potential therapeutic approach for ischemic stroke, improving functional recovery and alleviating secondary brain injury. Upon activation, resting endogenous NSPCs in the SVZ niche could proliferate and migrate toward the lesion and further differentiate into neurons to participate in reconstructing infract white matter. However, the detailed functions of ART still require further investigation before clinical trials. The present study indicated that ART could rescue penumbra damage and alleviate white matter injury, promoting neurogenesis and contributing to post-stroke functional recovery through FOXO3a/ p27 Kip1 pathway in MCAO mice (Fig. 6). These findings suggested that ART could be a novel therapeutic candidate for neurogenesis and functional reconstruction after stroke.
Data Availability Not applicable and the datasets used in this study were available from the corresponding author upon reasonable request.
The Ethics Committee of the Third Military Medical University approved all the animal work protocols (approval no. AMU-WEC20201153). The experiments were performed in concordance with Animal Research standards: Reporting of In Vivo Experiments (ARRIVE) guidelines.

Declarations
Consent to Participate Not applicable.

Consent for Publication
All the authors agreed to the publication.

Competing Interests
The authors declare no competing interests.