Functional Recovery Caused by Human Adipose Tissue Mesenchymal Stem Cell-Derived Extracellular Vesicles Administered 24 h after Stroke in Rats

Ischemic stroke is a major cause of death and disability, intensely demanding innovative and accessible therapeutic strategies. Approaches presenting a prolonged period for therapeutic intervention and new treatment administration routes are promising tools for stroke treatment. Here, we evaluated the potential neuroprotective properties of nasally administered human adipose tissue mesenchymal stem cell (hAT-MSC)-derived extracellular vesicles (EVs) obtained from healthy individuals who underwent liposuction. After a single intranasal EV (200 µg/kg) administered 24 h after a focal permanent ischemic stroke in rats, a higher number of EVs, improvement of the blood–brain barrier, and re-stabilization of vascularization were observed in the recoverable peri-infarct zone, as well as a significant decrease in infarct volume. In addition, EV treatment recovered long-term motor (front paws symmetry) and behavioral impairment (short- and long-term memory and anxiety-like behavior) induced by ischemic stroke. In line with these findings, our work highlights hAT-MSC-derived EVs as a promising therapeutic strategy for stroke.

The pathophysiology of ischemic stroke is characterized by blood flow obstruction in a restricted brain region, forming an infarct nucleus surrounded by an area known as the penumbra zone (known as the peri-infarct region in animal models) [8], which is an affected zone that can be recovered [9]. Reperfusion of the penumbra/peri-infarct Figure 1 shows the representative images of cell characterization. There is no single marker for hAT-MSCs; thus, they are characterized by immunophenotyping based on the presence (>70%) of cluster of differentiation (CD)90 and CD105 associated with the absence (<5%) of CD34 and CD45 [44]. Three sources of hAT-MSCs were used in this study, namely Cell (C) 0 (commercial, Figure 1A), C1 (donor 1, Figure 1B), and C2 (donor 2, Figure 1C), which were characterized by flow cytometry and fluorescence microscopy. In flow cytometry, the displacement of the fluorescence peak on the right side showed a positive value for the presence of protein markers. More than 70% of the C0, C1, and C2 were positive for CD90 and CD105 (Figure 1(Aa,Ba,Ca)), while only 0.3% of the analyzed cells were positive for CD45 and CD34 (Figure 1(Ab,Bb,Cb)), which is a characteristic of hAT-MSCs. The cells were also characterized by fluorescence microscopy using the same markers ( Figure 1(A(c,d),B(c,d),C(c,d))).
namely Cell (C) 0 (commercial, Figure 1A), C1 (donor 1, Figure 1B), and C2 (donor 2, Fi ure 1C), which were characterized by flow cytometry and fluorescence microscopy. flow cytometry, the displacement of the fluorescence peak on the right side showed a po itive value for the presence of protein markers. More than 70% of the C0, C1, and C2 we positive for CD90 and CD105 (Figure 1Aa,Ba,Ca), while only 0.3% of the analyzed ce were positive for CD45 and CD34 (Figure 1Ab,Bb,Cb), which is a characteristic of hA MSCs. The cells were also characterized by fluorescence microscopy using the same mar ers ( Figure 1A(c,d),B(c,d),C(c,d)).  Figure 2 shows the three sources of EVs used in this study, namely, EV0 obtaine from C0 (Figure 2A), EV1 obtained from C1 ( Figure 2B), and EV2 obtained from C2 (Figu 2C), which were characterized using four protocols. EVs were detected inside the cells b confocal microscopy through the presence of CD63 and CD81 EV markers (Figu 2Aa,Ba,Ca). Released EVs were analyzed using flow cytometry; more than 90% of the EV presented CD63 and CD81 [45,46]. Histograms are presented for EV0 (Figure 2Ab Figure 2 shows the three sources of EVs used in this study, namely, EV0 obtained from C0 (Figure 2A), EV1 obtained from C1 ( Figure 2B), and EV2 obtained from C2 ( Figure 2C), which were characterized using four protocols. EVs were detected inside the cells by confocal microscopy through the presence of CD63 and CD81 EV markers ( Figure 2(Aa,Ba,Ca)). Released EVs were analyzed using flow cytometry; more than 90% of the EVs presented CD63 and CD81 [45,46]. Histograms are presented for EV0 ( Figure 2  Cc)); the EVs suspension consisted only of vesicles with a cylindrical morphology and electron-dense membranes. The Zetasizer instrument was used to measure an average diameter of 140 nm, with a PDI average of 0.3 for EV0 (red), EV1 (green), and EV2 (blue) ( Figure 2D).

Front Paws Symmetry
Effect of Stroke on Paws Symmetry Figure 3 shows that all animals were subjected to the cylinder test (CT) 24 h before stroke (day-1), and only animals with~100% front paws symmetry before surgery were included in the study. Seventy-two hours after stroke, all ischemic (ISC) groups (treated and untreated) presented a mean symmetry of~30%.
Dose curve effect of EV0 treatment on front paws symmetry. Figure 3a shows the dose curve used to determine the lowest dose for EV treatment. Intranasal administration of EV0 or vehicle was performed 24 h after stroke with the naive group as control. Animals treated with EV0 (100, 200, or 300 µg/kg) showed a time-and dose-dependent better improvement in front paws symmetry (from day seven after treatment) compared to the ischemic untreated animals; however, only animals treated with 200 or 300 µg/kg EVs attained symmetry scores similar to those of the naive group (total recovery). Thus, a dose of 200 µg/kg was used in further experiments.
Effect of 200µg/kg EV treatment on front paws symmetry: Figure 3b shows that EV0, EV1, and EV2 treatments applied 24 h after a stroke caused better improved timedependent recovery of front paws symmetry compared to the ISC group (from day seven after treatment). Total recovery was attained on day 28 for all EVs. Supplementary Figure  S1 shows individual animal symmetry values of each EV treatment effect compared to the ISC group.

Infarct Volume and BBB Permeability
(e) Quantification of Evans blue in the brain. Statistical analysis using unpaired t test. Data are reported as mean ± SD. *** p < 0.001 compared to naive and sham groups, # p < 0.05 compared to ISC group. (f) Albumin levels in CSF. Statistical analysis using unpaired t test. Data are reported as the mean ± SD. ** p < 0.01 and *** p < 0.001 compared to naive and sham groups, # p < 0.05 compared to the ISC group. Figure 5 shows the distribution of EV0 (200 µg/kg) in the cerebral cortex evaluated by images acquired in three positions (Figure 5a). The regions examined were the supplementary motor cortex (M2) and somatosensory (SS) regions, both in the ipsi-and contra-lateral hemispheres. Remarkably, there was no homogeneous distribution of EVs throughout the brain. In the ISC group, the M2-ipsilateral (M2-I) and M2-contralateral (M2-C) regions presented a greater number of vesicles compared to the naive animals and to the ISC SS-ipsilateral (SS-I) and ISC SS-contralateral (SS-C) regions (Figure 5b). Figure 5c-h shows representative images of these findings. mean ± SD. *** p < 0.001 compared to naive and sham groups, # p < 0.05 compared to ISC group. (f) Albumin levels in CSF. Statistical analysis using unpaired t test. Data are reported as the mean ± SD. ** p < 0.01 and *** p < 0.001 compared to naive and sham groups, # p < 0.05 compared to the ISC group. Figure 5 shows the distribution of EV0 (200 µ g/kg) in the cerebral cortex evaluated by images acquired in three positions ( Figure 5a). The regions examined were the supplementary motor cortex (M2) and somatosensory (SS) regions, both in the ipsi-and contralateral hemispheres. Remarkably, there was no homogeneous distribution of EVs throughout the brain. In the ISC group, the M2-ipsilateral (M2-I) and M2-contralateral (M2-C) regions presented a greater number of vesicles compared to the naive animals and to the ISC SS-ipsilateral (SS-I) and ISC SS-contralateral (SS-C) regions ( Figure 5b). Figure  5c-h shows representative images of these findings.  Figure 6 shows that the same animals were subjected to all three sequential OFT sessions on days 7, 21, and 42 after PBS or EV treatment, using the naive group as control for evaluating memory of habituation to novelty. All groups presented normal short-term memory (evaluated only in the first session). Only the ISC group presented impairment of long-term memory (evaluated by comparing the first with the third session), an effect reversed with all EV treatment.  Figure 6 shows that the same animals were subjected to all three sequential OFT sessions on days 7, 21, and 42 after PBS or EV treatment, using the naive group as control for evaluating memory of habituation to novelty. All groups presented normal short-term memory (evaluated only in the first session). Only the ISC group presented impairment of long-term memory (evaluated by comparing the first with the third session), an effect reversed with all EV treatment.  Figure 7 shows that the same animals were subjected to all three sequential sessions on days 7, 21, and 42 after PBS or EV (EV0, EV1, and EV2) treatment, using the naive group as control. Stroke impaired both short-and long-term NORT memory, and effects abolished by EV treatment. , and (f) ISC + EV2 (n = 16). Statistical analysis using 2-way ANOVA, followed by the Tukey's multiple comparisons test. Data are reported as mean ± SEM. ** p < 0.01, *** p < 0.001, **** p < 0.0001 comparing the 1st min with the 5th min only in the first session; # p < 0.05, ## p < 0.01, comparing the 1st min of the first session with the 1st min of the second/third or third sessions. NS = not statistically significant. Figure 7 shows that the same animals were subjected to all three sequential sessions on days 7, 21, and 42 after PBS or EV (EV0, EV1, and EV2) treatment, using the naive group as control. Stroke impaired both short-and long-term NORT memory, and effects abolished by EV treatment.

Elevated Plus-Maze Task (EPMT)
The task was performed only on the seventh day after treatment with EV0 or EV1. ISC animals spent less time in the open arms compared to the other groups, indicating a stroke-induced anxiogenic-like effect, which was abolished by EV0 and EV1 treatment ( Figure 8).

Brain Angiogenesis
The number of branches and total length of blood vessels were evaluated in M2 and SS brain regions through images acquired in three positions (+2.20, +0.2, and −1.88 mm A.P. to Bregma) (Figure 9a).  (c,f,i) 3 successive short-term memory test sessions; (d,g,j) 3 successive long-term memory test sessions. Unpaired t test with a theoretical average of 50%. Data reported as mean ± SD; * p < 0.05, ** p < 0.01, and *** p < 0.001, compared with the theoretical average of 50%.

Elevated Plus-Maze Task (EPMT)
The task was performed only on the seventh day after treatment with EV0 or EV1. ISC animals spent less time in the open arms compared to the other groups, indicating a stroke-induced anxiogenic-like effect, which was abolished by EV0 and EV1 treatment ( Figure 8).  (c,f,i) 3 successive short-term memory test sessions; (d,g,j) 3 successive long-term memory test sessions. Unpaired t test with a theoretical average of 50%. Data reported as mean ± SD; * p < 0.05, ** p < 0.01, and *** p < 0.001, compared with the theoretical average of 50%. (c,f,i) 3 successive short-term memory test sessions; (d,g,j) 3 successive long-term memory test sessions. Unpaired t test with a theoretical average of 50%. Data reported as mean ± SD; * p < 0.05, ** p < 0.01, and *** p < 0.001, compared with the theoretical average of 50%.
A.P. to Bregma) (Figure 9a). Figure 9b,c show the number of branches and total length of blood vessels specifically in M2-ipsi-(I) and M2-contralateral (C) regions. Stroke decreased the number of branches and total length in both regions; however, EV2 treatment abolished the decrease specifically in the M2-I region. Figure 9d-i shows representative images of blood vessels in the M2-I and M2-C regions. In the SS regions, there was no difference of blood vessel parameters among the groups (See Supplementary Material: Figure S2).

Discussion
Since ischemic stroke is a significant cause of morbidity and mortality [1], research to develop new therapeutic strategies is constantly needed [15,20,47,48]. Scientific data from stroke patients have shown that behavioral and motor impairments are dependent on the infarct core and the penumbra zone in which reversible cell damage may be recovered by endogenous brain reactivity. The intensity and speed of the penumbra zone recovery may determine the size of the core and functional recovery after stroke [8]. Thus, innovative (h) ISC M2-C; (i) ISC + EV2 M2-C. Statistical analysis using the unpaired t test. Data are reported as mean ± SD. * p < 0.05, ** p < 0.01, **** p < 0.0001, compared to naive group; ### p < 0.001 compared to ISC group (n = 3). Scale bars = 20 µm.

Discussion
Since ischemic stroke is a significant cause of morbidity and mortality [1], research to develop new therapeutic strategies is constantly needed [15,20,47,48]. Scientific data from stroke patients have shown that behavioral and motor impairments are dependent on the infarct core and the penumbra zone in which reversible cell damage may be recovered by endogenous brain reactivity. The intensity and speed of the penumbra zone recovery may determine the size of the core and functional recovery after stroke [8]. Thus, innovative experimental therapeutic proposals, including the administration of EVs derived from MSCs, target this peri-infarct region, aiming to improve the recovery process [49] and, consequently, functional recovery [40]. Our research groups, and others, have already shown that, in the ischemic stroke rat model used here, the core and peri-infarct regions are located in the prefrontal cortex and hippocampus [17][18][19][20][21]50], which are brain structures involved in neuromotor and behavioral performance [51] evaluated in this study.
EVs are important mediators of communication between cells, and it has been identified that native brain cell-released EVs present tropism to injured regions [52][53][54]. Additionally, it has already been shown that intranasally administrated MSC-derived EVs were identified in brain cells [41,42], presenting tropism to injured brain regions [49]. In this study, we demonstrated that a single dose of EVs derived from hAT-MSCs intranasally administrated 24 h after brain insult resulted in a higher number of EVs in the peri-infarct regions, resulting in a decrease in infarct volume, alteration of the BBB permeability, and new vascularization. Notably, EVs administration reversed the impairments caused by brain insult on front paws symmetry, short-and long-term memory in OFT and NORT, and anxiety-like behavior. Thus, EV treatment contributed to the recovery of the peri-infarct brain region and simultaneously reversed the impairment in motor and behavioral performance, pointing to a potential role of hAT-MSC-derived EVs in functional recovery after stroke.
Stimulation of angiogenesis, as observed by EV treatment, has been shown to improve neurological and motor function in animal stroke models [55], an effect currently acknowledged as an outcome of EV transfer of proteins, mRNAs, and miRNAs to endothelial cells [33,56] and regulating protein expression [57]. Additionally, BBB impairment in ischemic stroke, as observed here, has also been documented [58][59][60], but its involvement in EVs therapeutic strategies, as indicated in this study, has not been previously reported.
Here, we demonstrate that intranasal hAT-MSC-derived EVs administered 24 h after stroke promoted long-term neuroprotective effects, offering a remarkably therapeutic window. Although the three EVs show protective properties, findings are not comparable since EV0, EV1, and EV2 are from different sources (2 healthy individuals and commercial hAT-MSCs). Together, these findings indicate that hAT-MSC-derived EVs are a promising potential therapeutic strategy for patients with focal permanent ischemic stroke.

hAT-MSCs: Sources, Culture and Characterization
The cells were obtained from commercial and human sources.

Patient-Derived hAT-MSCs
Cells were obtained from the subcutaneous adipose tissue of two (32-year-old and 34-year-old) women who underwent abdominal liposuction at the Hospital de Clínicas in Porto Alegre, RS, Brazil. Informed consent was obtained from all subjects involved in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by Research and Graduate Group (Grupo de Pesquisa e Pós-Graduação: GPPG 2018-0374) and Research Ethics Committee (Comitê de Ética em Pesquisa CAEE: 94521618.4.0000.5327) of the Experimental Research Center at Hospital de Clínicas de Porto Alegre. Fresh adipose tissue was washed with PBS buffer, minced, and digested for 1 h in 0.1% collagenase at 37 • C. The digestion process was stopped by the addition of DMEM containing 20% FBS (Cripion). The digested suspension was filtered through a 70 µm nylon mesh cell filter to retain tissue debris. The filtered suspension was centrifuged at 400× g for 5 min. The stromal vascular fraction (pellet) was resus-pended in DMEM + 20% FBS (Cripion) medium and cultured in a culture flask containing 25 cm 2 (TPP -Techno Plastic Products, Trasadingen, CH) at 37 • C in a humidified 5% CO 2 atmosphere. After 24 h, non-adherent cells were gently removed [61]. When adherent cells reached 80% confluence (passage 0: P0), confluent cells (hAT-MSCs) were detached with 0.25% trypsin/1 mM ethylenediaminetetraacetic acid (EDTA) (Sigma-Aldrich) and plated in flasks at a density of 1.5 × 10 4 cells/75 cm 2 (TPP) (passage1: P1). Cells were then cultured in DMEM (Sigma-Aldrich) containing 10% FBS (Cripion), 100 units/mL penicillin (Gibco/Thermo Fisher Scientific), 100 µg/mL streptomycin (Gibco/Thermo Fisher Scientific), gentamicin 50 mg/L (Sigma-Aldrich), fungizone (2.5 mg) (Sigma-Aldrich), and L-glutamine 4 mM (Gibco/Thermo Fisher Scientific). These cells were named cell 1patient 1 (C1) and cell 2-patient 2 (C2). C1 and C2 were expanded under the same conditions described above and used only from the 4th to 8th passage [42,61]. C0, C1, and C2 hAT-MSCs were characterized by immunofluorescence using flow cytometry and confocal microscopy.

Flow Cytometry
hAT-MSCs were centrifuged at 400× g for 5 min at room temperature, and the cell pellet was re-suspended in DMEM + 10% FBS and counted in a Neubauer chamber. Briefly, the cells were incubated with antibodies at a concentration of 1:50 for 4 h at 37 • C. The cell suspensions were centrifuged at 400× g for 5 min at room temperature, and cell pellets were resuspended in 200 µL of PBS. Ten thousand events were analyzed using flow cytometry (FACSCalibur™-BD Biosciense, Franklin Lakes, NJ, USA) [44]. Cells in passage 4 (P4) were characterized as hAT-MSCs by the presence of CD: CD34 (FITC conjugate mouse antihuman) (BD Biosciense), CD45 (FITC conjugate mouse anti-human) (Invitrogen, Waltham, MA, USA), CD90 (P.E. conjugate mouse anti-human) (BD Biosciense), and CD105 (R-PE conjugate mouse anti-human) (Invitrogen).

Confocal Microscopy
An aliquot of 1 × 10 4 hAT-MSCs was placed on a slide and analyzed by immunofluorescence. Cells were maintained under culture conditions for 72 h to allow adherence to coverslips. Cells were then incubated for 4 h at 37 • C with the same antibodies used for cytometry: CD34, CD45, CD90, and CD105, at a ratio of 1:500. The negative control was prepared by incubating only the secondary antibodies, Alexa Fluor 555 and 488 (Invitrogen). Cells on coverslips were gently washed with PBS (four times) to remove excess antibodies and then fixed in 4% PFA for 2 h. Following fixation, the cells were gently washed again with PBS and then fixed with Fluoromount (Sigma-Aldrich) onto a slide for further analysis. Images were acquired using an 8-bit grayscale confocal laser scanning microscope (Olympus FV1000, Shinjuku, Tokyo, Japan). Approximately 10 × 15 sections with 0.7 µm thick confocal were captured parallel to the coverslip (XY sections) using a ×20 objective. Z-stack reconstruction and analysis were conducted using ImageJ software (http://rsb.info.nih.gov/ij/, accessed on June 2018).
For EV isolation, the collected medium was centrifuged (3 times) at 4 • C: (400× g for 15 min, then 2000× g for 15 min, and then 10,000× g for 30 min). The supernatants were filtered through a 0.22 µm membrane. The isolation was completed by centrifugation (100,000× g at 4 • C for 2 h). The supernatant was discarded, PBS was used to wash the pellet containing EVs, and the cell suspension was centrifuged at 100,000× g at 4 • C for 2 h. This centrifugation protocol (speed and time) was shown to preserve EVs in the pellet while preventing contamination with stress-related biological constituents [46,[65][66][67]. Finally, the pellet was resuspended in 100 µL of PBS and stored at −20 • C [67]. Protein content was measured using a bicinchoninic acid (BCA) assay (Thermo Fisher Scientific) [61]. The vesicles isolated from C0, C1, and C2 cells were named EV0, EV1, and EV2, respectively.

EVs Characterization
EVs were characterized by flow cytometry through the identification of membrane proteins CD63 and CD81 [45,46,[65][66][67]. First, EVs were incubated with magnetic beads (Thermo Fisher Scientific, Invitrogen™), coated with primary antibody CD63 (Thermo Fisher Scientific) and CD81 (Thermo Fisher Scientific) for 18 h at 4 • C under gentle stirring. For each measurement, 10 µL of 1 mg/mL EV suspension was applied. To remove excess beads, immediately after incubation, EVs were washed with PBS, 2 mL of PBS was added for 5 min, and then the tube was placed in a magnet (to remove the beads) for 1 min, and the supernatant was discarded. Then, CD63, clone: MEM-259 (Invitrogen) and CD81, clone JS-81 (BD Pharmingen™, San Diego, CA, USA) antibodies (without granules) were added to the solution containing the EVs + magnetic beads. After 1 h of incubation, the EVs were gently washed by placing the tube on a magnet for 1 min, and the supernatant was discarded. We added 2 mL of PBS (to remove excess antibody) for 5 min and again placed the tube on a magnet for 1 min and discarded the supernatant. Finally, the EVs were resuspended in 200 µL PBS for analysis. Ten thousand events were analyzed by flow cytometry.
We used photon correlation spectroscopy to measure the particle size and polydispersity index (PDI). The EV suspension derived from hAT-MSCs (50 µL) at 1 mg/mL was diluted in 1 mL of PBS. All analyses were performed in triplicate using a Malvern Nano-ZS90 ® (Malvern Instruments, Marvin City, UK) at 25 • C.

EVs Measurement
Transmission electron microscopy (TEM) analysis, using a direct examination technique, was used to evaluate the diameter of EVs [49]. EV suspension (10 µL) and 1 mg/mL of protein were pipetted in aliquots into a grid covered with a carbon film (formvar/carbon) and dried at room temperature. Uranyl (Merck KGaA, Darmstadt, DE, USA) was used as a contrast agent. The sample was analyzed by TEM at 120 Kv in Microscopy and Microanalysis Center-UFRGS (JEM 1200 Exll-JEOL, Tokyo, Japan).

EVs Labeling
EVs were labeled with the red fluorescent membrane dye PKH26 (Sigma-Aldrich). In brief, the EV-containing PBS solution was centrifuged at 100,000× g for 2 h at 4 • C, and the pellet was suspended with the diluent of the fluorescent kit. Filtered PKH26 (4 mM) and EVs (200 µg/mL) were mixed at a ratio of 1:1 for 5 min, followed by the addition of 5% BSA. To remove excess dye, the EVs were washed three times, and then 5 mL of PBS was added and centrifuged at 100,000× g for 2 h at 4 • C and the supernatant discarded. In the last centrifugation step, the stained EV pellet was suspended in PBS (0.5 mL). The solution was filtered through a 0.2 µm membrane filter to remove dye aggregates [42].

Animals
Adult (90-120 days old) male Wistar rats weighing 350-400 g were maintained under controlled light (12/12 h light/dark cycle) at 22 • C ± 2 • C, with water and food ad libitum. All procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals and the Brazilian Society for Neuroscience and Behavior recommendations for animal studies. This study was approved by the Ethics Committee for the Use of Animals at the Universidade Federal do Rio Grande do Sul (project identification: 31888). A schematic of the procedure is shown in Figure 10.
Adult (90-120 days old) male Wistar rats weighing 350-400 g were maintained under controlled light (12/12 h light/dark cycle) at 22 °C ± 2 °C, with water and food ad libitum. All procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals and the Brazilian Society for Neuroscience and Behavior recommendations for animal studies. This study was approved by the Ethics Committee for the Use of Animals at the Universidade Federal do Rio Grande do Sul (project identification: 31888). A schematic of the procedure is shown in Figure 10. Focal permanent ischemia and sham procedures: Anesthetized animals (under ketamine hydrochloride: 90 mg/kg i.p. and xylazine hydrochloride:10 mg/kg i.p.) were placed into a stereotaxic apparatus. After skin incision, the skull was exposed, and a craniotomy was performed by exposing the left frontoparietal cortex (+2 to −6 mm A.P. and −2 to −4 mm M.L. from the bregma). A focal permanent ischemic lesion was induced by thermocoagulation of the motor and sensorimotor pial vessels. Blood vessels were thermocoagulated by placing a hot probe near the dura mater for 2 min until a red-brown color indicated complete thermo-coagulation. Soon after, the skin was sutured, and the animals were placed on a heating pad at 37 °C until full recovery from anesthesia. Animals from the sham group were subjected to craniotomy as described above. Animals were randomly allocated to three treatment groups: sham, ischemic (ISC), and ischemic treated Focal permanent ischemia and sham procedures: Anesthetized animals (under ketamine hydrochloride: 90 mg/kg i.p. and xylazine hydrochloride:10 mg/kg i.p.) were placed into a stereotaxic apparatus. After skin incision, the skull was exposed, and a craniotomy was performed by exposing the left frontoparietal cortex (+2 to −6 mm A.P. and −2 to −4 mm M.L. from the bregma). A focal permanent ischemic lesion was induced by thermocoagulation of the motor and sensorimotor pial vessels. Blood vessels were thermocoagulated by placing a hot probe near the dura mater for 2 min until a red-brown color indicated complete thermo-coagulation. Soon after, the skin was sutured, and the animals were placed on a heating pad at 37 • C until full recovery from anesthesia. Animals from the sham group were subjected to craniotomy as described above. Animals were randomly allocated to three treatment groups: sham, ischemic (ISC), and ischemic treated with EVs (ISC + EV). Our stroke model of focal cerebral permanent ischemia represents a protocol to investigate brain lesions with good reproducibility and low mortality [17][18][19][20][21]50,68,69].

Extracellular Fluorescent Vesicle (EV) Detection in the Rat Brain
The distribution of fluorescent EVs (PKH26-Sigma-Aldrich) in rat brains (naive and ISC + EV1) was analyzed 18 h after intranasal administration, as previously reported [42,49].
Anesthetized animals (ketamine hydrochloride: 90 mg/kg, 450 µL/kg i.p. and xylazine hydrochloride: 10 mg/kg, 300 µL/kg i.p.) were transcardially perfused with PBS using a peristaltic pump, followed by perfusion with 4% PFA (both 10 mL/min, 100 mL). Immediately, brains were dissected, immersed in 4% PFA (pH 7.4), and stored for a maximum of 7 days at 4 • C. Coronal brain sections (20 µm thick) were obtained using a vibratome (Leica, Wetzlar, Germany) at +2.20, 0,20, and −1,88 mm of Bregma (PAXINUS online Rat Brain Atlas: http://labs.gaidi.ca/rat-brain-atlas/ (accessed on January 2020). Brain slices were incubated for 5 min in the dark with 1 µg/mL Hoechst 33342 solution dye (Sigma-Aldrich) to detect cell nuclei. The slices were washed with PBS (four times), and the slices were fixed with the fluoro mount (Sigma-Aldrich). Slice images for counting EVs were acquired using an 8-bit grayscale confocal laser scanning microscope (Olympus FV1000). Approximately 10-15 sections with 0.7 µm thick confocal were captured parallel to the coverslip (XY sections) using a ×60 objective. Z-stack reconstructions and analyses to count the vesicles in the brain tissue were conducted using ImageJ Free software (National Institutes of Health, Bethesda, MD, USA), and background noise was removed using the "subtract background" tool. Images were converted to binary masks using the default threshold option, and vesicles were counted with the "analyze particles" tool (size = 0.05-0.90 µm). These settings were programmed into a macro and used for all the analyzed images (http://rsb.info.nih.gov/ij/ (accessed on June 2018)).

Short-Term Evaluation of Infarct Volume
For evaluating infarct volume, naive, sham, ISC, ISC + EV0, and ISC + EV2 animals were sedated 48 h after treatment (O 2 flow rate of 0.8-1.0 Â mL/min with isoflurane levels of 2.5-3.0%), decapitated, and the brain removed. Coronal sections of the whole brain were sliced at 2 mm, and slices were immersed in 2% 2, 3, 5-Triphenyl-tetrazolium chloride (TTC) (Sigma-Aldrich) for 30 min at 37 • C. After incubation, the slices were dipped in 4% buffered paraformaldehyde (pH 7.4) for 24 h. The infarct area was evaluated as an area devoid of red staining. The infarct volume was measured using the ImageJ software [21].

Brain Angiogenesis
After 42 days of treatment, animals from the naive, ISC, and ISC + EV2 groups were anesthetized and received intracardiac injection of 50 mg/mL (500 µL) fluorescein isothiocyanate-dextran amine (Merck KGaA) to label blood vessels. After 5 min, rat brains were excised, immediately fixed in 4% PFA, and cut into 30 µm coronal slices in a vibratome. Images were acquired using a fluorescence microscope (Nikon, Tokyo, Japan). The images were taken from the ipsilateral and contralateral sides in the secondary motor cortex (M2) and somatosensory (SS) regions using the following coordinates: +2.20, 0.2, and −1.88 mm A.P. to Bregma (PAXINUS online Rat Brain Atlas: http://labs.gaidi.ca/rat-brain-atlas/ (accessed on January 2020). Blood vessel parameters, such as the total length (sum of the length of segments and isolated elements) and the number of branches, were quantified using the Angiogenesis Analyzer Plugin (Gilles Carpentier Research) ImageJ software [72] (https://imagej.nih.gov/ij/ accessed on June 2019) [73,74].

BBB Permeability
Evans blue penetration into brain parenchyma: Naive, sham, ISC, and ISC + EV2 animals were anesthetized (ketamine hydrochloride: 90 mg/kg, 450 µL/kg i.p. and xylazine hydrochloride 10 mg/kg, 300 µL/kg i.p.) 48 h after treatment, and 3 mL/kg of 2% Evans blue (EB) (Sigma-Aldrich, San Luis, MI, USA) solution in saline was administered through the gingival artery (Supplementary information Figure S1). After 1 h, the animals were subjected to cardiac perfusion using a peristaltic pump (10 mL/min, with PBS, 100 mL). The animals were decapitated, and whole brain images were obtained. The brains were then sliced (coronal sections) at 2 mm and images obtained. The slices from each brain were macerated and homogenized in 2.5 mL of PBS and vortexed for 2 min to measure the amount of EB in the brain parenchyma. For protein precipitation, 2.5 mL of 50% trichloroacetic acid was added to the homogenate, incubated for 12 h at 50 • C, and centrifuged at 14,000× g for 10 min. The concentration of the blue color was measured in the supernatant using a spectrophotometer at a wavelength of 620 nm. EB dye was expressed in micrograms per gram of brain tissue [75,76].

Behavioral Tasks
A group of animals was submitted to the cylinder task, open field task, and novel object recognition task; another group of animals was submitted to the cylinder task and elevated plus-maze task. The cylinder task, which evaluates neurologic dysfunction (front paws symmetry), was performed before surgery; the presence of~100% basal symmetry was used as inclusion criterion in the study.

Cylinder Task (CT)
The CT can evaluate the motor symmetry of the front paws. The apparatus consists of a transparent glass cylinder 20 cm in diameter and 30 cm in height (20 raising movements were counted). All animals (naive, naïve + EV0, sham, ISC, and ISC + EVs groups) were firstly submitted to this task 24 h before surgery to verify basal symmetry. This task evaluated how the rats raised their bodies in contact with their paws on the cylinder wall. The ipsilateral (to the lesion), contralateral, or both front paws preferences were counted in a blinded analysis. The asymmetry of each animal was calculated using the following formula: asymmetry = (% of ipsilateral use = ipsilateral paw use/sum ipsilateral + contralateral + use of both paws) − (% of contralateral paw use/sum of ipsilateral and contralateral paws). The performance was recorded using ANY-Maze software version 6.3 (Stoelting Co., Wood Dale, IL, USA). The asymmetry percentage was converted into a symmetry percentage [50]. The same group of animals were submitted to CT 24 h before surgery and on the 3rd, 7th, 14th, 21st, 28th, 35th, and 42nd day after EV treatment. At the end of each task, the apparatus was cleaned using 70% ethanol solution.

Open Field Task (OFT)
This task evaluates habituation to novelty (assessing short-and long-term memory through exploratory activity) and locomotor activity [79,80]. The arena consisted of a black cage measuring 50 × 50 × 50 cm. The individual sessions lasted 10 min. The animals (naive, naïve + EV0, ISC, and ISC + EVs groups) were submitted to the task on days 7, 21, and 42 after EV treatment. Short-term memory was evaluated considering the decrease in locomotion during the first 5 min of the 1st session (only on the 7th day). Long-term memory was evaluated considering the decrease in locomotion during the first minute through successive sessions (from the 1st to the 3rd session). At the end of each session, the apparatus was cleaned with 70% ethanol solution. The task was recorded and analyzed using ANY-Maze software.

Novel Object Recognition Task (NORT)
Behavioral sessions lasting 10 min were performed on days 7, 21, and 42 after EV treatment. At 90 min after the OFT session, object recognition (OR) short-and long-term memories were evaluated [81]. The animals were individually placed on the periphery of the arena for further exploration. Two identical familiar objects (FOs) were placed in the arena, and the animals could explore them for 10 min (training session). Sniffing and touching objects were considered exploratory behaviors. The animals were then removed from the arena, and 90 min after the training session, each animal was placed back into the arena to evaluate short-term memory (first test session). One of the two FOs used in the training session was replaced by a new distinct object (NO). Long-term memory was evaluated 24 h after the training session, when the animals were placed back in the arena with the same FO used in the training session and the first test session (short-term memory); however, the same NO was displaced to a different position. In all sessions, the time spent exploring the objects was recorded using ANY-maze software. The results are expressed as a percentage of the time spent exploring each object. Animals that recognized the novel object (short-term memory) or its new position (long-term memory) explored more than 50% of the total exploration time of both objects. At the end of each session, the apparatus was cleaned using 70% ethanol solution.

Elevated Plus-Maze Task (EPMT)
This task is widely used to study anxiety-like behavior [82]. The apparatus had two open arms (50 cm long × 10 cm wide) and two closed arms (50 cm long × 10 cm wide × 40 cm high), separated by a square central platform (5 × 5 cm). The apparatus was placed 70 cm above the floor. The animals (naive, naïve + EV0, ISC, ISC + EV0, and ISC + EV1) were habituated in a red-light room for 1 h before starting the task. The percentage of time spent in the open and closed arms was assessed. Anxiety-like behavior was considered as the increase of time spent in the closed arms. Each animal was submitted to this task once on the 7th day after treatment with EVs. ANY-maze software was used to record behavioral performance for 5 min. For baseline, we used the results of naive group performance. At the end of each session, the equipment was cleaned with 70% alcohol.

Statistical Analyses
Brain lesion size and the number of brain vesicles showed a non-homogeneous distribution (Shapiro-Wilk test p < 0.05); thus, statistical analyses of both experiments were performed using the Mann-Whitney test, and the data were reported as median and interquartile range.
The results of BBB integrity, angiogenesis, OFT short-term memory, and EPMT were evaluated using unpaired t tests. Two-way RM ANOVA was applied for CT, followed by Sidak's multiple comparison test. OFT long-term memory was evaluated by two-way ANOVA, followed by Sidak's multiple comparison test. Unpaired t tests were used for the NORT, with a theoretical average of 50%. Comparison between EVs (EV0, EV1, and EV2) were not performed due to differences in EVs sources (2 donors and one commercial). In the parametric distribution, the data were presented as mean ± SD. All analyses were performed using GraphPad Prism version 6.0 (San Diego, CA, USA). All participants who donated adipose tissue for cell isolation signed a free and informed consent. This informed consent describes the entire objective of the research and that these data can be published in a scientific journal, maintaining the confidentiality of the participants' data (anonymous donation).

Conflicts of Interest:
The authors declare no potential conflicts of interest concerning the research, authorship, and/or publication of this article.