Male Sprague-Dawley (SD) rats (n=35; weight, 300 g; age, 9–10 weeks), were purchased from Slykingda Experimental Animal [Hunan, China; permit no. scxk (Xiang) 2016–0002]. Pregnant SD rats (n=3; weight 300 g; age, 9–10 weeks), were purchased from Tianqin Biotech [permit no. scxk (Xiang) 2016–0217]. All animal experiments and animal care were conducted in accordance with the criteria of the Laboratory Animals Welfare Act, the Guide for the Care and Use of Laboratory Animals provided by the Institutional Animal Care and Use Committee of Nanchang University. All experimental protocols for the use of animals were approved by the Animal Care and Use Committee of Nanchang University (Nanchang, China). All rats were housed under pathogen-free conditions at 30–70% humidity and 26°C and had access to standard rodent food and water ad libitum and maintained under a 12-h light/dark cycle. Experiments were performed on rats between 7 and 10 weeks of age. Animals were euthanized after completion of the experiments and prior to tissue collection by CO₂ asphyxiation at a flow rate of 20% cage volume displacement/minute (5 l/min). Death after exposure to CO₂ was confirmed based on careful assessment of the rats for cardiac arrest.

TUNEL assay kit (cat. no. C1088) was purchased from Beyotime Institute of Biotechnology. Rabbit antibodies against CD31 (cat. no. bs-20321R; 1:1,000) and GSK-3β (cat. no. bs-0028R; 1:1,000) were obtained from BIOSS; rabbit antibody against VEGF (cat. no. AF5109; 1:1,000) was obtained from Affinity; mouse monoclonal antibodies against β-actin (cat. no. TA-09; 1:2,000); HRP-conjugated goat anti-mouse IgG (H + L; cat. no. ZB-2305; 1:2,000) and HRP-conjugated goat anti-rabbit IgG (H + L; cat. no. ZB-2301; 1:2,000) were purchased from Zhongshan Golden Bridge Biotechnology Co., Ltd (OriGene Technologies, Inc.); rabbit polyclonal anti-phosphorylated (p)-GSK-3β (cat. no. AF2016; 1:500) was purchased from Affinity Biosciences, Ltd.; PVDF membrane (cat. no. IPVH00010) was purchased from MilliporeSigma; SuperSignal West Pico Chemiluminescent Substrate (cat. no. 34077) was obtained from Thermo Fisher Scientific, Inc.; Ultrasensitive Chemiluminescence Imaging system (ChemiDoc XRS⁺) and CFX Connect Real-Time PCR Detection system were purchased from Bio-Rad Laboratories, Inc. Ultrapure RNA Extraction kit (cat. no. CW0581M) was purchased from CWBIO; HiScript II Q RT SuperMix for qPCR (cat. no. R223-01) was obtained from Vazyme; and Universal SYBR Green qPCR Master Mix was purchased from Applied Biosystems; Thermo Fisher Scientific, Inc.

EPC isolation was performed as reported previously (19). Briefly, 3-day-old neonatal SD rats (n=3) from the pregnant females were sacrificed by decapitation and sterilized by soaking in 75% ethanol for 5 min. The tibia and femur were isolated, and the attached muscles were removed. The tibia and femur were washed with PBS and the bone marrow was washed into a Petri dish. The bone marrow was repeatedly pipetted to form a single cell suspension, which was then carefully added to the surface of 4 ml mixture of Ficoll, hydroxyethyl starch 550 and meglumine diatrizoate (20) and centrifuged at 500 × g for 20 min at 25–26°C. The cells in the buffy coat fractions were collected, diluted with EBM-2 medium (cat. no. CC-3156; Lonza Group, Ltd.) and pelleted at 500 × g at 25–26°C for 5 min. Cells were resuspended in EBM-2 medium and cultured in 2% CO₂ at 37°C. The cells were then cultured in serum-free EBM-2 medium in a 2% CO₂ incubator at 37°C for 48 h, collected and stored at −80°C until exosome extraction. All operations were performed in laminar hoods to avoid microbial contamination.

EPCs (10⁴ cells/ml) were inoculated onto a microscope cover glass and cultured in EBM-2 medium in 2% CO₂ at 37°C until cells reached 90% confluency. The slides were washed with PBS three times (3 min each), fixed at 25–26°C with 4% paraformaldehyde for 15 min and permeated with 0.5% Triton X-100 (prepared in PBS) at room temperature for 20 min. The slides were then soaked in PBS for 5 min for three times (3 min each) at 25–26°C. Cells were blocked with 5% BSA (CoWin Biosciences) at 37°C for 30 min. Diluted primary rabbit anti-coagulation factor VIII antibody (cat. no. bs-2974R; BIOSS; 1:200) was added and the slides were incubated at 4°C overnight. The slides were subsequently incubated with Cy3-conjugated goat anti-rabbit IgG secondary antibody (1:200; cat no. S0011; Affinity Biosciences) at 37°C for 30 min. The nuclei were stained with DAPI at 25–26°C for 1 h and the slides were examined under a fluorescence microscope.

EPCs were rapidly thawed at 37°C and the supernatant was centrifuged at 2,000 × g for 30 min at 4°C. The supernatant was centrifuged again at 12,000 × g for 45 min at 4°C to remove larger vesicles. The supernatant was then filtered through a membrane (0.45 µm pore size) and pelleted by centrifuging at 11,0000 × g for 70 min at 4°C. The pellet was resuspended with 10 ml precooled 1X PBS. The exosome suspension was injected into a NanoFCM N30E nanoflow detector (Malvern Instruments, Ltd.) to determine the size distribution (diameter and number).

EPCs were fixed in 2.5% glutaraldehyde at 25–26°C for 1 h. After washing in pre-cooled PBS, the EPCs were dehydrated using ethanol and acetone, soaked in embedding solution overnight at room temperature and embedded in epoxy resin. Embedded cell blocks were cut into ultrathin sections (50-nm) and stained with 2% uranyl acetate for 30 min at 25–26°C, and washed with water five times (10 sec each time). Then, the sections were stained with 1% lead citrate for 15 min at 25–26°C and washed five times (10 sec each time) before TEM at 80 kV.

A classical suture method was used to establish IR models (21). Briefly, rats were anesthetized by intraperitoneal injection of ketamine 100 mg/kg (Shanghai Hengrui Pharmaceutical Co., Ltd.) and xylazine 10 mg/kg (Hubei Xinmingtai Chemical Co., Ltd.). A smooth incision was made along the middle line of the neck to separate bluntly the left sternocleidomastoid muscles and cervical muscles and to expose the right common carotid artery. The carotid artery was separated at the trident point to expose internal and external carotid arteries. The proximal ends of the right common carotid artery and external carotid artery were ligated, and the internal carotid artery was clamped with a vascular clip. A small incision was made on the right common carotid artery 1 cm away from the trigeminal nerve. A monofilament was introduced along the carotid artery into the brain to block the blood flow of the middle cerebral artery. After 2 h embolization, the thread was withdrawn. Nerve function defect was evaluated 24 h after reperfusion and scored as follows: i) 0, no symptoms of nerve injury; ii) 1, incomplete extension of the left front paw; iii) 2, circling left; iv) 3, falling to the left; and v) 4, loss of consciousness and unable to walk autonomously (22).

The 35 male SD rats were randomly divided into five groups (n=7 rats/group): i) untreated rats (control); ii) sham operation without plus inserted (sham); iii) IR model rats injected with 50 µl PBS (model); iv) model rats injected with 50 µl EPC cell suspensions at 6×10⁶ cells/ml (model + EPC); v) and model rats injected with 50 µl exosome suspension at 0.6 µg/µl (model + exosome) once a day for 3 days. The dose of exosomes used was selected based on a previous study (23).

Three days before modelling, rats were anesthetized, fixed on a brain stereotaxic apparatus and a 1.5 cm longitudinal incision was made in the middle of the skin of the head. A skull drill was used to make a hole 0.22 mm posterior to and 10 mm to the right side of the bregma; care was taken to avoid damaging the dura. A total of 50 µl suspension (aforementioned) was administered using a microsyringe injector into the lateral ventricle below the surface of the skull. The needle was withdrawn slowly 5 min following injection, and the injection site was sterilized twice with iodophor and sutured. The mice were then placed on a thermal pad and reared in the cage when the animals were awake from the anesthesia.

Rats were anesthetized as aforementioned and perfused with 20 ml PBS into the brain. The brain was dissected to isolate the cerebellum, brain stem and olfactory bulb. The brain tissues were frozen at −20°C for 30 min and the cerebellum and olfactory bulb were removed. The remaining brain tissue was sectioned (2 mm thick) and stained in 2% TTC dye solution in the dark at 37°C for 15 min; during this period, the sections were turned over every 5 min. The infarcted area was gray-white, and the non-infarcted area was dark red.

H&E staining was conducted to examine the tissue damage as previously described (24). Briefly, the brain tissue was dehydrated in an ascending series of ethanol (70, 80, 90 and 100%) at 25–26°C for 5 min at each concentration and cleared with xylene. Dehydrated tissue was embedded in paraffin, sectioned (4-µm thick), dewaxed with xylene and rehydrated in a series of ethanol (100, 90, 80, 70, 50, 30 and 0%) at 25–26°C for 5 min at each concentration. The sections were stained with an aqueous hematoxylin solution at 25–26°C for 3 min, differentiated with hydrochloric acid for 15 sec, briefly washed with tap water at 25–26°C for 60 sec, and counterstained with eosin at 25–26°C for 3 min. Sections were washed in distilled water, dehydrated and cleared as previously described, then the sections were sealed and examined under a CX41 light microscope (Olympus Corporation) at ×200 magnification to observe pathological changes, including the number of nerve cells and glial cells.

A TUNEL assay was used to detect apoptotic cells in the brain tissues as described previously (25). Briefly, brain tissue sections (4-µm) were baked at 65°C for 2 h, rehydrated in a descending ethanol series and treated with proteinase K (50 µg/ml) for 30 min at 37°C. The sections were rinsed with PBS three times (5 min each) and incubated with TUNEL detection solution at 37°C in the dark for 1 h, according to the supplier's protocols. The slide was then incubated with DAPI at room temperature in the dark for 3 min; the excess DAPI was rinsed away with PBS and the slide was blotted dry with absorbent paper. The slides were sealed with anti-fluorescence quenching solution and observed under a CX41 fluorescence microscope (Olympus Corporation) in 10 fields of view.

Total RNA was extracted from the brain tissues using an RNA Extraction kit (Takara Bio, Inc.) according to the manufacturer's instructions. RNA concentrations were quantified using a Nanodrop spectrophotometer (NanoDrop Technologies; Thermo Fisher Scientific, Inc.) and subsequently reverse transcribed into cDNA using the High-Capacity cDNA Transcriptase Reverse kit (Applied Biosystems; Thermo Fisher Scientific, Inc.) according to manufacturer's protocol. qPCR was conducted using the Universal SYBR Green qPCR Master Mix (Applied Biosystems; Thermo Fisher Scientific, Inc.) on a CFX96 Real-Time PCR Detection System (Bio-Rad Laboratories, Inc.) using the primers listed in Table I. Relative mRNA expression levels were determined using the 2^−ΔΔCq method after normalization with β-actin as an internal reference (26). qPCR was carried out in a total volume of 15 µl containing 1 µl of diluted and pre-amplified cDNA, 10 µl Universal SYBR Green qPCR Master Mix and 1.5 µl of each forward and reverse primer. The thermocycling conditions were as follows: Initial denaturation at 95°C for 10 min; followed by 40 cycles of 95°C for 15 sec and 57°C for 60 sec.

Brain sections (4-µm) were fixed at 25–26°C in 4% paraformaldehyde for 10–15 min and rinsed three times with PBS (3 min each). The cells were cleared with 0.5% Triton X-100 (in PBS) at room temperature for 20 min and washed with PBS three times (5 min each). After blocking with 5% BSA at 37°C for 30 min, anti-CD31 antibody (1:1,000) was added and the plates were incubated overnight at 4°C. The plates were then immersed in PBS three times (3 min each) and incubated with Cy3-conjugated goat anti-rabbit IgG (1:200; cat. no. CW0159S; CoWin Biosciences) at 25–26°C for 1 h. Subsequently, the slides were incubated with an anti-VEGF antibody (1:1,000) at 25–26°C for 30 min and then incubated with diluted Alexa Fluor 488-conjugated goat anti-rabbit IgG (1:200; cat. no. ZF-0511; Zhongshan Golden Bridge Biotechnology Co., Ltd; OriGene Technologies) at 37°C for 45 min and counterstained with DAPI at 25–26°C in the dark for 5 min. Images were captured using a fluorescence microscope (Olympus Corporation).

Brain tissues (0.2 g) were lysed with RIPA buffer (Beijing Solarbio Science & Technology Co., Ltd.) containing protease inhibitors cocktail and quantitated using a BCA kit (CoWin Biosciences) according to the manufacturer's instructions. After denaturing by boiling at 100°C for 5 min, 50 µg protein was separated by 10% SDS-PAGE, transferred to PVDF membranes, blocked with 5% non-fat milk in 1X TBS-0.1% Tween-20 buffer for 4 h at room temperature and then detected by incubation with the following primary antibodies (at the aformentioned dilutions) at 4°C overnight: Mouse monoclonal anti-β-actin, rabbit polyclonal anti-Wnt3α, rabbit polyclonal anti-Gsk-3β and rabbit polyclonal anti-p-Gsk-3β. Subsequently, the membranes were incubated with HRP-conjugated goat anti-mouse IgG or HRP-conjugated goat anti-rabbit IgG secondary antibodies (at the aformentioned dilutions) at 25–26°C for 1 h. Protein bands were visualized using the SuperSignal West Pico Chemiluminescent Substrate (cat. no. 34077; Thermo Fisher Scientific, USA). Densitometric analysis was conducted using Quantity One software (version v4.6.6; Bio-Rad Laboratories, Inc.) using β-actin as the internal control.

Data are expressed as the mean ± standard error of the mean obtained from at least three independent experiments. Statistical comparisons between groups were assessed using one-way ANOVA with Tukey's post hoc tests. Ordinal data obtained for nerve defect scoring were analyzed using the Kruskal-Wallis test followed by Dunn's post hoc tests. Statistical analysis was performed using SPSS 21.0 software (IBM Corp.). P<0.05 was considered to indicate a statistically significant difference.

Immunofluorescence results demonstrated that isolated EPCs had red fluorescence with a wavelength 640 nm emitted from factor VIII (Fig. 1A). TEM and nanoflow measurements confirmed that the isolated exosomes exhibited the cup-like shape with double membranes (Fig. 1B) and were in the expected size range (30–299 nm), with the majority of exosomes being 60–80 nm in diameter (Fig. 2).

Results from TTC staining revealed that the infarcted area increased significantly after IR modeling compared with the control and sham groups (Fig. 3A). EPC and exosome treatments significantly reduced the infarcted area compared with the untreated model group (Fig. 3A). Similarly, IR modelling significantly increased nerve defects compared with the control (Fig. 3B); the EPC and exosome treatments significantly reduced the defect score (Fig. 3B), and the improvement was more notable with exosome than with EPC (P<0.05; Fig. 3B).

H&E staining revealed that in control and sham rats, the boundary of the cortex and gray matter of the brain was clear without edema and necrosis; the nerve cells were arranged orderly and evenly, the cell membrane was intact and there was a clear nucleus and nucleolus (Fig. 4). In model group, the nerve fibers were slightly necrotic and swollen, the number of nerve cells was reduced and glial cells were proliferated. In the EPC-treated rats, the brain tissue showed mild liquefaction and degeneration, mild edema in the stroma and decreased number of nerve cells which were distributed less evenly; glial cells were proliferated. In the exosome-treated rats, the distribution of nerve cells was more uniform, the degeneration and necrosis of cells were less intensive, and the proliferation of glial cells was remarkable (Fig. 4).

The TUNEL assay results revealed that, compared with the control, the number of apoptotic cells was significantly increased after IR modelling (P<0.05; Fig. 5). Compared with the untreated model group, apoptosis was significantly decreased following EPC and exosome treatments (both P<0.05). Compared with EPC treatment, model rats treated with exosomes exhibited a significant reduction in apoptosis (P<0.05).

qPCR results revealed that the mRNA expression levels of CD31 and VEGF were significantly increased after IR modelling compared with the control group (P<0.05; Fig. 6); the expressions were further upregulated following EPC and exosome treatments compared with the untreated model group (both P<0.05). Similarly, immunofluorescence assay results demonstrated that the protein expression levels of CD31 and VEGF were significantly increased after IR modelling (P<0.05) compared with the control and sham groups, and they were further upregulated following EPC and exosome treatments (P<0.05), particularly with exosomes (P<0.05) (Fig. 7) compared with the model group.

Western blot analysis revealed that the protein expression levels of Wnt3a, GSK-3β and p-GSK-3β were significantly increased after IR modelling (P<0.05) compared with the control and sham groups, and were significantly downregulated in model rats treated with EPCs or exosomes compared with the control and sham groups; no significant difference in the ratio of p-GSK-3β to GSK-3β expression levels were detected (P>0.05) compared with the control and sham groups (Fig. 8). Exosome treatment resulted in a more marked downregulation compared with EPC treatment (P<0.05).