Exosomes Derived from CXCR4-Overexpressing BMSC Promoted Activation of Microvascular Endothelial Cells in Cerebral Ischemia/Reperfusion Injury

Background Ischemic stroke is a severe acute cerebrovascular disease which can be improved with neuroprotective therapies at an early stage. However, due to the lack of effective neuroprotective drugs, most stroke patients have varying degrees of long-term disability. In the present study, we investigated the role of exosomes derived from CXCR4-overexpressing BMSCs in restoring vascular function and neural repair after ischemic cerebral infarction. Methods BMSCs were transfected with lentivirus encoded by CXCR4 (BMSCCXCR4). Exosomes derived from BMSCCXCR4 (ExoCXCR4) were isolated and characterized by transmission electron microscopy and dynamic light scattering. Western blot and qPCR were used to analyze the expression of CXCR4 in BMSCs and exosomes. The acute middle cerebral artery occlusion (MCAO) model was prepared, ExoCXCR4 were injected into the rats, and behavioral changes were analyzed. The role of ExoCXCR4 in promoting the proliferation and tube formation for angiogenesis and protecting brain endothelial cells was determined in vitro. Results Compared with the control groups, the ExoCXCR4 group showed a significantly lower mNSS score at 7 d, 14 d, and 21 d after ischemia/reperfusion (P < 0.05). The bEnd.3 cells in the ExoCXCR4 group have stronger proliferation ability than other groups (P < 0.05), while the CXCR4 inhibitor can reduce this effect. Exosomes control (ExoCon) can significantly promote the migration of bEnd.3 cells (P < 0.05), while there was no significant difference between the ExoCXCR4 and ExoCon groups (P > 0.05). ExoCXCR4 can further promote the proliferation and tube formation for the angiogenesis of the endothelium compared with ExoCon group (P < 0.05). In addition, cobalt chloride (COCl2) can increase the expression of β-catenin and Wnt-3, while ExoCon can reduce the expression of these proteins (P < 0.05). ExoCXCR4 can further attenuate the activation of Wnt-3a/β-catenin pathway (P < 0.05). Conclusions In ischemia/reperfusion injury, ExoCXCR4 promoted the proliferation and tube formation of microvascular endothelial cells and play an antiapoptotic role via the Wnt-3a/β-catenin pathway.


Introduction
Stroke is an acute cerebrovascular disease caused by local cerebral blood circulation disorder, which is the second cause of global human death [1]. According to the report published by the World Health Organization in 2017, about 6.24 million people die of stroke every year [2]. Among the surviving stroke patients, over 70% of them have varying degrees of long-term disability [3]. Stroke includes ischemic and hemorrhagic stroke, and the former is the most common subtype, accounting for 87% of all cerebrovascular accidents [4]. Currently, intravenous alteplase and mechanical thrombectomy are recommended effective treatments for patients with acute ischemic stroke. However, these two treatments have many limitations, especially the narrow time window, which largely limits their clinical use [5]. Thus, it is necessary to find alternative therapeutics for patients with acute ischemic stroke. The brain microvascular endothelium is a thin layer of connected and anchorage-dependent cells, which constitutes the interface between the bloodstream and the deformable solid vascular wall [6]. This vascular endothelium is a highly active metabolic system that synthesizes various vascular regulatory factors to adjust the microcirculation of the cerebral tissue [7]. The brain microvascular endothelium cells are now recognized as a highly active metabolic system that synthesizes various materials to nourish nerves and regulate vasomotor function [8]. As is well known, effective collateral circulation can protect brain tissue after cerebral ischemia by increasing blood perfusion in the ischemic penumbra (IP) [9]. It has been demonstrated that microvascular integrity and cerebrovascular angiogenesis may stabilize brain perfusion thereby promoting neuronal survival, brain plasticity, and neurologic recovery [10][11][12][13]. Thus, enhancing angiogenesis in ischemic brain tissue is considered to be an emerging opportunity for potential therapeutic strategies.
Bone marrow mesenchymal stem cells (BMSCs) are able to secrete a variety of trophic factors promoting cell repair and growth [14]. Many studies indicated that stem cells can inhibit inflammation and promote angiogenesis in the ischemic area. Therefore, stem cell therapy is considered to be one of the most promising treatments for ischemic stroke [15,16]. CXC motif chemokine receptor type 4 (CXCR4) belongs to G-proteincoupled receptor superfamily, which is involved in the homing of a variety of cells. The ligand of CXCR4 is stromal cell-derived factor 1 (SDF-1 or CXCL12) belonging to CXC chemokine family. The expression of CXCL12 is significantly increased in ischemic brain tissue [17] because of the activation of hypoxia-inducible factor-1 (HIF-1). Previous studies have shown that plasma CXCL12 will increase in poststroke patients, which indicated that the concentration gradient of CXCL12 will form in the blood when it is released from the ischemic area [18,19]. To verify the angiogenic roles of CXCL12 in vivo, our previous study utilized the chicken chorioallantoic membrane (CAM) assay. We found that CXCL12 in doses of 50 and 100 ng/ml induced neovessel formation, which indicated that CXCL12 enhanced the capacity for angiogenesis in vivo [20]. CXCL12 can bind to the CXCR4 on CXCR4-expressing cells, including brain microvascular endothelial cells (EC). This promotes the migration of cerebral microvascular endothelial cells to ischemic tissues improving the brain tissue repair. Vessel sprouting demands migration and polarization of ECs in response to cytokines [21]. Our previous study indicated that CXCR4 may be involved in the migration and polarization of human umbilical vein endothelial cells in the stripe assay [20]. Collectively, these results are the foundation of the present study to investigate the role of CXCR4 in promoting angiogenesis and protecting brain endothelial cells.
Exosomes are small cell membrane vesicles with a diameter of 30-100 nm; the contents are lipids, proteins, RNA, miRNAs, and mRNAs [22,23]. Exosomes are carriers of intercellular signal transmission, playing an important role in intercellular communication. The improvement of neural function depends on the entire neurovascular unit, including neurons, astrocytes, vascular endothelial cells, basement membrane, and extracellular matrix. A signal transduction network between cells plays an important role in tissue reconstruction after cerebral ischemia [24]. Increasing evidence is demonstrating that exosomes of mesenchymal stem cells can promote vascular remodeling in the ischemic zone after cerebral ischemia [25]. However, whether the exosomes deprived from CXCR4-overexpressing BMSCs play a role in the angiogenesis is still unclear.
In the present study, we hypothesized that exosomes derived from BMSCs could effectively regulate cell survival, promote angiogenesis, protect nerve cells, and improve stroke outcome in the acute stage of stroke, through upregulating CXCR4 in exosomes of BMSCs. We aim to investigate the role of exosomes derived from CXCR4-overexpressing BMSCs in restoring vascular function and neural repair after ischemic cerebral infarction and to clarify the mechanism of these exosomes in promoting angiogenesis and protecting endothelial cells, so as to provide new ideas for the treatments and further mechanism research of stroke.

Exosome Isolation and Identification.
After 24 h of incubation in a serum-free medium, the BMSC supernatant was collected and subjected to 300g for 10 min, 3000g for 20 min, 10,000g for 30 min, and 100,000g for 2 h of ultracentrifugation at 4°C. Exosomes were obtained and stored at -80°C. The exosomes were resuspended in PBS and were spread evenly over the copper grid. Then, they were negatively stained with the phosphotungstic acid for 10 min and were observed using a transmission electron microscopy.
The number and size of exosomes were assessed using the NanoSight NS300 system (Malvern Instruments, UK). The supernatant was diluted in PBS, and 1 ml of the solution was used for NanoSight analysis. Each sample was analyzed for 5 times, 10 s for each time. Western blot was used to detect the expression of Alix, TSG101, and CD9 proteins of the exosomes.
2.6. Cell Uptake of Exosomes. The exosomes were labeled with PHK26 (Sigma, USA) and were incubated with bEnd.3 in DMEM supplemented with 10% FBS for 48 hours. Then, they were washed with PBS to stop the cell absorption and fixed in 4% paraformaldehyde. Finally, the outcome of cell absorption was observed using the Olympus BX41 microscope equipped with a CCD camera.

Middle Cerebral Artery Occlusion in Rats.
Twenty-four male adult SD rats weighing 250-270 g were divided into 4 groups with 6 rats in each group: the sham-operated group (sham group), the acute middle cerebral artery occlusion model (MCAO) rats injected with PBS group (control group), the MCAO rats injected with exosomes derived from BMSC Con group (Exo group), and the MCAO rats injected with exosomes derived from BMSC CXCR4 group (Exo CXCR4 group). Rats were anesthetized with 1% pentobarbital intraperitoneally and then received an operation according to the modified Zea Longa method [26]. Briefly, a 6-0 coated with silicon nylon monofilament suture (Beijing Cinontech Co., Ltd., Beijing, China) was introduced into external carotid artery lumen and then gently advanced into the internal carotid artery in order to block the origin of the middle cerebral artery. For the rats in the sham-operated group, the suture was inserted into external carotid artery lumen 3 Neural Plasticity but not advanced into the internal carotid artery. After 90 min occlusion, the suture was withdrawn to allow the reperfusion. A 5-point scale [26] was used to assess the neurological deficiency score, and the rats with a score between 1 and 3 were considered as a success of MCAO.
2.9. Exosome Transplantation. Exosome transplantation was conducted at 24 h after the establishment of MCAO. The injection was performed as described previously [27]. At a speed of 5 μl/min, 100 μg Exo, Exo CXCR4 , or the same volume of PBS was injected into the lateral ventricle of the affected hemisphere (X = −0:5 mm, anterior-posterior; Y = 1:5 mm, medial-lateral; and Z = −3:5 mm, dorsal-ventral). The observation of vocalization, seizures, hemiplegia, and body weight was continued after surgery.
2.10. Triphenyl Tetrazolium Chloride (TTC) Staining. Rats were anesthetized with 2% pentobarbital intraperitoneally at 3 d after ischemia/reperfusion (I/R), and the brains were removed after profusion with NS. A brain slicer was used to coronally section the brains at 2 mm intervals from the frontal pole. The slices were incubated with 1% TTC solution for 15 minutes at 37°C in the dark and then were immersed with 4% paraformaldehyde.
2.11. Modified Neurological Severity Score. Rats were examined with modified neurological severity score (mNSS) at 1 d, 3 d, 7 d, 14 d, and 21 d after I/R. This evaluation was performed by a blinded tester. mNSS consists of motor (muscle status, abnormal movement), sensory (visual, tactile, and proprioceptive), reflex, and balance test. Scores ranged from 0 to a maximum of 14. The extent of neurological deficits was assessed as follows: 1-5 scores, mild deficits; 6-9 scores, moderate deficits, and 10-14 scores, severe deficits.
2.12. Adhesive Removal Test. The adhesive remove test was used to assess the integration of motor and sensory of rats at 1 d, 3 d, 7 d, 14 d, and 21 d after I/R. This evaluation was performed by a blinded tester. For each rat, an adhesive paper with the size of 1 × 1 cm was pasted on the surface of its forelimb plantar, and the time of removing the adhesive paper was recorded.
2.13. CCK8 Detects Cell Proliferation. The bEnd.3 cells were seeded in a 96-well plate at a density of 1 × 10 4 cells per well, and 5 wells were set as one group. In the experiment of exosomes promoting the proliferation of bEnd.3 cells, the bEnd.3 cells were divided into 5 groups according to the concentrations of exosome (0, 5, 10, 15, and 20 μg/ml); in the experiment of cobalt chloride inhibiting the proliferation of bEnd.3 cells, the bEnd.3 cells were divided into 5 groups according to the concentrations of cobalt chloride (0, 100, 200, 400, and 600 μmol/l). Fresh medium containing 10% CCK8 solution was added into the wells and cultured in a 5% CO 2 -saturated humidity incubator at 37°C for 0.5 to 4 h. The absorbance value (OD value) of bEnd.3 cells at a wavelength of 450 nm was detected by a microplate reader.
2.14. Edu Staining Assay Detects Cell Proliferation. The bEnd.3 cells were divided into 4 groups: the bEnd.3 cells cul-tured with exosomes derived from BMSC Con group (Exo Con group); the bEnd.3 cells cultured with exosomes derived from BMSC CXCR4 group (Exo CXCR4 group); the bEnd.3 cells cultured with exosomes derived from BMSC CXCR4 and AMD3100 group (Exo CXCR4 + AMD3100 group), and the bEnd.3 cell control group. 50 μmol/l Edu was added to the medium of bEnd.3 cells in each group and then cultured in a 5% CO 2 incubator at 37°C. All cell nuclei were stained blue, while the proliferating cell nuclei were also stained red, which were quantified by the fluorescence microscope.
2.15. Cell Migration Assay. Cell migration was measured by a Transwell cell chamber. Conditioned medium with 15 μg/ml Exo Con or Exo CXCR4 was added to the lower chamber, while 1 × 10 5 bEnd.3 cells were cultured in the upper chamber. After 24 h, the cells were fixed with formaldehyde and then stained with crystal violet (Sigma, USA). Migratory bEnd.3 cells were quantified by a microscope.
2.16. Tube Formation Assay. The bEnd.3 cells were incubated with a medium containing exosome or not for 24 h were collected. Cells in each group were seeded in 48-well plates (5 × 10 3 cells per well) coated with matrix glue, and each group was established in three holes. After 6 h, the vascular tubular morphogenesis was observed in at 100x magnification by the inverted optical microscope. The number of vascular lumen structures in 5 random visual field was recorded by ImageJ software. 2.18. Statistical Analysis. SPSS 20.0 software (IBM Corporation, NY, USA) and GraphPad Prism 7.0 (GraphPad Software Inc., CA, USA) were used for data analyses; all data were expressed as the mean ± standard deviation (SD). All sample data conformed to the normal distribution by the Kolmogorov-Smirnov test (P > 0:05). Differences between multiple groups were analyzed by one-way analysis of variance (ANOVA) test, and then, multiple comparisons were performed using Bonferroni post hoc test. Values of P < 0:05 are considered statistically significant. 4 Neural Plasticity

Identification of BMSCs and Exosomes.
BMSCs were isolated from SD rats, and the expression of immune markers in the third generation of them was identified by flow cytometry. As shown in Figure 1(a), the CD105 (94.5%), CD44 (98.6%), CD73 (91.7%), and CD90 (96.2%) were found to be positive, while the CD34 (5.10%) were found to be negative, which indicated the high purity of BMSCs.
The shape of the exosomes derived from BMSCs was a classic cup shape under transmission electron microscopy ( Figure 1(b)). NTA tracking was used to analyze the size of exosomes, which indicated that the diameter distribution ranged from 50 nm to 600 nm, with a peak at 100 nm (Figure 1(c)). In addition, the expression of exosome protein markers was detected by Western blot. As shown in Figure 1(d), CD9, TSG101, and Alix were expressed in the exosomes. When exosomes labeled with the fluorescent PKH26 were incubated with bEnd.3 cells, the recipient cells-bEnd.3 cells-exhibited high uptake efficiency, as demonstrated by a fluorescence microscopy (Figure 1(e)).

Increased Expression of CXCR4 in Exosomes Derived from
BMSCs Transfected with Lentivirus. The lentivirus conveying control vector and CXCR4-overexpressing vector were constructed and transfected into BMSCs. Compared with control groups, the protein and RNA expression of CXCR4 increased in the BMSCs which were transfected with lentivirus encoded by CXCR4 (Figures 2(a)-2(c)). As shown in Figures 2(d), the expression of Alix protein was similar between the Exo Con and Exo CXCR4 groups, which indicated the similar concentration of exosome in both groups. Compared with the Exo Con group, the expression of CXCR4 protein was significantly increased in the Exo CXCR4 group (P < 0:05, Figure 2(e)).

Injection of Exo CXCR4 Improved Neurobehavioral
Outcome in MCAO Rats. The neurobehavior outcomes of the rats were measured for day 1, day 3, day 7, day 14, and day 21 (Figure 3(a)). We measured the cerebral infarction volume of the rats for day 3 by TTC staining in both the sham and MCAO groups and found that the cerebral infarct volumes were significantly different among the 2 groups (n = 6, Figure 3(b)). As shown in Figure 3(c), compared with the control group and Exo Con group, the Exo CXCR4 group showed a significantly lower mNSS score at 14 d after I/R (P < 0:05, n = 6, Figure 3(c)). For the time of removing sticky papers, there was no significant difference among each groups (P > 0:05, n = 6, Figure 3(d)).

The
Role of Exo CXCR4 in the Proliferation, Migration, and Tube Formation of bEnd.3 Cells. Exo Con were added to bEnd.3 cells. It was found that the proliferation of bEnd.3 cells was enhanced with increasing exosome concentrations. The proliferative ability of bEnd.3 cells reached its peak at the exosome concentration of 15 μg/ml, which was used in the following experiments (Figure 4(b)). Then, bEnd.3 cells were cultured with Exo Con , Exo CXCR4 , or Exo CXCR4 plus AMD3100, and the Edu staining and CCK8 were used to detect the proliferation of bEnd.3 cells. As shown in Figures 4(a), 4(c), and 4(d), the bEnd.3 cells in the Exo CXCR4 group have stronger proliferation ability than other groups (P < 0:05), while the CXCR4 inhibitor (AMD3100) can reduce this effect. As for the migration of bEnd.3 cells, Exo Con can significantly promote the migration compared with the control group (P < 0:05), while there was no significant difference between the Exo CXCR4 and Exo Con groups (P > 0:05, Figures 5(a) and 5(b)). As indicated in Figures 5(c) and 5(d), compared with the control group, Exo Con can promote the angiogenesis of the endothelium (P < 0:05). Additionally, the Exo CXCR4 can further promote the angiogenesis of the endothelium compared with the Exo Con group (P < 0:05). The addition of AMD3100 will seriously affect the tube formation.

Exo CXCR4 Can Reduce Cell Apoptosis Caused by Cobalt
Chloride (COCl 2 ). Western blot was used to assess the 6 Neural Plasticity expression of the proapoptotic protein BAX and antiapoptotic proteins Bcl-2 and PARP. As shown in Figures 6(a) and 6(c), compared with the control group, COCl 2 can significant increase the expression of BAX and reduce the expression of Bcl-2 and PARP (P < 0:05). Compared with the COCl 2 group, the expression of Bcl-2 and PARP in the COCl 2 + Ex o Con group was higher, while the expression of BAX was lower (P < 0:05). In addition, Exo CXCR4 can further reduced cell apoptosis caused by COCL 2 . Finally, the potential mechanism of antiapoptosis effect of Exo CXCR4 was detected by Western blot. As shown in Figures 6(b), 6(d), and 6(e), COCl 2 can increase the expression of β-catenin and Wnt-3, while Exo Con can reduce the expression of these proteins. Additionally, Exo CXCR4 can fur-ther attenuate the activation of the Wnt-3a/β-catenin pathway. This indicated that the antiapoptosis effect of Exo CXCR4 was related to the Wnt-3a/β-catenin pathway.

Discussion
The present study takes advantage of the important role of the CXCL12/CXCR4 axis in the recovery of cerebral infarction. The expression of CXCL12 was significantly upregulated after cerebral ischemia, leading to chemotaxis of endothelial progenitor cells expressing CXCR4 [28]. Additionally, hours after cerebral infarction, CXCL12 expression was elevated in both astrocytes and endothelial cells [29]. Compared to CXCR7, another receptor for CXCL12, 7 Neural Plasticity CXCR4, is mostly expressed on the cell membrane [30]. Exosomes are vesicles that are packaged by the cell membrane for information transmission [31]. Therefore, it was hypothesized that by increasing CXCR4 on the cell membranes, the amount of CXCR4 on exosomes could be increased. In the present study, BMSCs were genetically modified to overexpress CXCR4 with the help of lentiviral transfection and the exosomes secreted by BMSCs were isolated. According to the results of Western blot and qPCR, CXCR4 was detectable on the Exo CXCR4 group, while very little CXCR4 expression was found on Exo Con from BMSCs transfected with no-load virus. Additionally, the recipient cells, bEnd.3 cells, exhibited high uptake efficiency of exosomes as demonstrated by a fluorescence microscopy. These results provided foundation for the research of exosomes in vivo and vitro.
In in vivo experiments, TTC staining was firstly used to assess the cerebral infarction volume of the rats for day 3 to make sure the success of the conduction of MCAO model. Then, the mNSS score and adhesive removal test were used to assess the recovery of the MCAO rats. The results of these tests shown that the injection of exosomes can accelerate the recovery of the MCAO rats, and the rats in the Exo CXCR4 group recovered the fastest. This indicated that CXCR4 might promote the positive effect of exosomes in ischemia reperfusion injury.
In in vitro experiments, Exo CXCR4 were able to promote the proliferation and tube formation of brain microvascular endothelial cells. These effects were all offset by AMD3100, a CXCR4 inhibitor. This evidence does not only prove that BMSC exosomes can promote the angiogenesis in brain tissues after ischemia but also indicates that CXCR4 can improve this effect. However, Exo CXCR4 did not promote cell migration better than ordinary BMSC exosomes. This may be because the taken-up exosomes did not completely transfer the CXCR4 to the cell membrane. Besides, the CXCR4 content transferred by exosomes is obviously lower than the CXCR4 content of the cell surface itself, so the total amount of CXCR4 on the cell surface did not change significantly.
BMSCs have a significant inhibitory effect on inflammation. Previous studies have injected BMSCs into animals with central nervous system damage and found that BMSCs can

Neural Plasticity
survive and migrate to damaged tissues and ultimately improve neural function [15,32]. The CXCL12/CXCR4 axis plays an important role in the treatment of ischemic stroke with transplanting BMSCs [33]. Previous studies have shown that expression of CXCL12 in astrocytes is upregulated within 7 days after cerebral ischemia in newborn mice [29]. CXCR4 is expressed on the surface of BMSCs, which may be an important reason that helps BMSCs migrate to the ischemic area. However, most intravascularly injected BMSCs migrate to the lungs, probably due to insufficient expression of CXCR4 in BMSCs [34,35]. Therefore, many studies have attempted injection of BMSCs with high expression of CXCR4 or stereotactic injection of BMSCs [36,37]. Different from BMSCs, exosomes were small with low immunogenicity. It is not easy for exosomes to block blood vessels and easy to store [25,38]. They are now increasingly used for the exploration of various diseases. Exosomes have been introduced into regenerative medicine as an alternative to BMSCs [39]. However, how to efficiently deliver exosomes to the ischemic area remains unclear. Similar to BMSCs, there are only a few CXCR4 on the surface of exosomes. Most of the exosomes circulate in the body during intravenous injection and their utilization efficiency is uncertain [40]. Therefore, in the present study, exosomes were derived from BMSCs which were transfected with lentivirus encoded by CXCR4. In vivo experiments determined that functional recovery of rats in the Exo CXCR4 group was significantly improved compared with rats injected with ordinary exosomes.

Neural Plasticity
After blood clot formation during ischemic stroke, endothelial cells undergo apoptosis due to lack of nutrients, which will destroy the blood-brain barrier, cause brain tissue edema, and aggravate disease progression [41,42]. The promotion of new vessels in the ischemic region is important for improving the outcome of cerebral infarction. Data have shown that in the 21 days after ischemia, new blood vessels are formed at the edge of the ischemic area, which gradually extends to the ischemic center through budding [9]. Previous studies have shown that CXCL12/CXCR4 can promote the stabilization of endothelial cells and improve the outcome of atherosclerotic diseases by antiapoptotic in vivo protein [43]. COCl 2 can cause hypoxia in cells, leading to autophagy and apoptosis of cells [44]. In the present study, COCl 2 can significantly increase the expression of proapoptotic protein BAX and reduce the expression of antiapoptotic proteins Bcl-2 and PARP, while Exo Con can reverse this effect caused by COCl 2 . Exo CXCR4 can further reduce cell apoptosis caused by COCL 2 . Additionally, COCl 2 can increase the expression of β-catenin and Wnt-3, while Exo Con can reduce the expression of these proteins. Exo CXCR4 can further attenuate the activation of the Wnt-3a/β-catenin pathway. Thus, the present study shows that Exo CXCR4 can promote the proliferation and tube formation of microvascular endothelial cells and plays an antiapoptotic role via the Wnt-3a/β-catenin pathway, which is consistent with the previous results.
However, this study mainly investigated the effect and mechanism of exosomes derived from BMSCs in promoting  10 Neural Plasticity angiogenesis in ischemia/reperfusion injury, without further research into the specific components of BMSC exosome by which promoted angiogenesis. This is the limitation in this study as well as our future research directions.

Conclusion
In ischemia/reperfusion injury, Exo CXCR4 can promote the proliferation and tube formation of microvascular endothelial cells and play an antiapoptotic role via the Wnt-3a/βcatenin pathway.

Data Availability
The raw data supporting the conclusions of this manuscript will be made available by the authors, without undue reservation, to any qualified researcher.

Conflicts of Interest
The authors declare that there is no conflict of interests regarding the publication of this paper.