Extracellular Vesicles Derived from Neural Progenitor Cells––a Preclinical Evaluation for Stroke Treatment in Mice

Stem cells such as mesenchymal stem cells (MSCs) enhance neurological recovery in preclinical stroke models by secreting extracellular vesicles (EVs). Since previous reports have focused on the application of MSC-EVs only, the role of the most suitable host cell for EV enrichment and preclinical stroke treatment remains elusive. The present study aimed to evaluate the therapeutic potential of EVs derived from neural progenitor cells (NPCs) following experimental stroke. Using the PEG technique, EVs were enriched and characterized by electron microscopy, proteomics, rt-PCR, nanosight tracking analysis, and Western blotting. Different dosages of NPC-EVs displaying a characteristic profile in size, shape, cargo protein, and non-coding RNA contents were incubated in the presence of cerebral organoids exposed to oxygen-glucose deprivation (OGD), significantly reducing cell injury when compared with control organoids. Systemic administration of NPC-EVs in male C57BL6 mice following experimental ischemia enhanced neurological recovery and neuroregeneration for as long as 3 months. Interestingly, the therapeutic impact of such NPC-EVs was found to be not inferior to MSC-EVs. Flow cytometric analyses of blood and brain samples 7 days post-stroke demonstrated increased blood concentrations of B and T lymphocytes after NPC-EV delivery, without affecting cerebral cell counts. Likewise, a biodistribution analysis after systemic delivery of NPC-EVs revealed the majority of NPC-EVs to be found in extracranial organs such as the liver and the lung. This proof-of-concept study supports the idea of EVs being a general concept of stem cell–induced neuroprotection under stroke conditions, where EVs contribute to reverting the peripheral post-stroke immunosuppression. Electronic supplementary material The online version of this article (10.1007/s12975-020-00814-z) contains supplementary material, which is available to authorized users.

Abstract: Stem cells such as mesenchymal stem cells (MSCs) enhance neurological recovery in preclinical stroke models by secreting extracellular vesicles (EVs). Since previous reports have focused on the application of MSC-EVs only, the role of the most suitable host cell for EV enrichment and preclinical stroke treatment remains elusive. The present study aimed to evaluate the therapeutic potential of EVs derived from neural progenitor cells (NPCs) following experimental stroke. Using the PEG technique, EVs were enriched and characterized by electron microscopy, proteomics, rt-PCR, nanosight tracking analysis, and Western blotting. Different dosages of NPC-EVs displaying a characteristic profile in size, shape, protein cargo, and non-coding RNA contents were incubated in the presence of cerebral organoids exposed to oxygenglucose-deprivation (OGD), significantly reducing cell injury when compared to control organoids. Systemic administration of NPC-EVs in male C57BL6 mice following experimental ischemia enhanced neurological recovery and neuroregeneration for as long as three months. Interestingly, the therapeutic impact of such NPC-EVs was found to be not inferior to MSC-EVs. Flow cytometric analyses of blood and brain samples seven days post-stroke demonstrated increased blood concentrations of B and T lymphocytes after NPC-EV delivery, without affecting cerebral cell counts. Likewise, a biodistribution analysis after systemic delivery of NPC-EVs revealed the majority of NPC-EVs to be found in extracranial organs such as the liver and the lung. This proofof-concept study supports the idea of EVs being a general concept of stem cellinduced neuroprotection under stroke conditions, where EVs contribute to reverting the peripheral post-stroke immunosuppression.
Dear Professor Zhang, Thank you for the encouraging response from your side. We have made the corrections as suggested by the two reviewers. Plus, we have still found some additional minor mistakes, which we have now corrected, too.
Thank you for considering our work for publication in Translational Stroke Research.
Best regards from Germany, Different dosages of NPC-EVs displaying a characteristic profile in size, shape, protein cargo, and non-coding RNA contents were incubated in the presence of cerebral organoids exposed to oxygen-glucose-deprivation (OGD), significantly reducing cell injury when compared to control organoids. Systemic administration of NPC-EVs in male C57BL6 mice following experimental ischemia enhanced neurological recovery and neuroregeneration for as long as three months. Interestingly, the therapeutic impact of such NPC-EVs was found to be not inferior to MSC-EVs. Flow cytometric analyses of blood and brain samples seven days poststroke demonstrated increased blood concentrations of B and T lymphocytes after NPC-EV delivery, without affecting cerebral cell counts. Likewise, a biodistribution analysis after systemic delivery of NPC-EVs revealed the majority of NPC-EVs to be found in extracranial organs such as the liver and the lung. This proof-of-concept study supports the idea of EVs being a general concept of stem cell-induced neuroprotection under stroke conditions, where EVs contribute to reverting the peripheral post-stroke immunosuppression.

Introduction
The systemic transplantation of stem cells such as mesenchymal stem cells (MSCs) and neural progenitor cells (NPCs), promotes neurological recovery, angiogenesis, and neurogenesis in animal models of cerebral ischemia [1][2][3][4][5][6][7]. The majority of these grafted cells, however, are trapped in extracerebral organs and do not integrate into existing neural networks. Rather, transplanted stem cells display both low survival rates and poor differentiation rates within the ischemic milieu [8][9][10][11], suggesting an indirect mode of action by which post-stroke neurological recovery is achieved.
It has been demonstrated that stem cell-derived conditioned medium induces similar effects in various disease models when compared to stem cell transplantation itself [12,13].
MSCs and other cell types secrete neurotrophic factors such as EGF and VEGF, thus inducing both neuroprotection and neuroregeneration [14][15][16]. Recent research, however, has questioned the hypothesis that these beneficial factors are the sole biological key mediator of stem cell-induced brain protection against cerebral ischemia. Instead, bilayer structured vesicles, termed extracellular vesicles (EVs), secreted from eukaryotic cells, including MSCs, have been found to be critical players in the aforementioned process. These EVs have also been detected in conditioned medium derived from stem cells [17,18], further supporting the idea that EVs are biological mediators of stem cell-induced actions under conditions of cerebral ischemia.
EVs are a heterogeneous group of vesicles ranging in size from 30 nm to 1000 nm.
They contain a defined set of cargo, which depends on the characteristics of the source cell [19].

Primary culture of MSCs and NPCs
P1 newborn mice were anesthetized using CO2 euthanasia. The brain was removed, and the subventricular zone (SVZ) was dissected in cold PBS under microscopic control. NPCs. Aliquots of 500 µl each were stored at -80°C until usage. The MSC-EV isolation was done after the third MSC passage. MSCs were cultured overnight with an FBS-free culture medium and after 24 h, the cell supernatants were collected. EVs were obtained from these supernatants by using the same protocol as for the NPC-EV PEG isolation method.
Since the optimal enrichment procedure for NPC-EVs is still a matter for debate [19], we applied differential centrifugation, i.e., ultracentrifugation only, for some experiments. As such, some NPC-EV samples were treated with ultracentrifugation as described before without the application of PEG 6,000. In brief, the cell culture medium was filtered through 220 nm filters to remove cell debris and apoptotic bodies followed by a 2 h ultracentrifugation procedure at 110,000 g.

Nanoparticle tracking analyses (NTA)
For both size determination and quantification of enriched NPC-EVs, an NTA was performed using the Nanosight platform (NanoSight LM10, Malvern Panalytical, Kassel, Germany). As shown previously [34], 1:1,000 water-diluted samples were measured in duplicate, and 400 µl of the diluted sample was injected into the measurement chamber. Each sample was measured three times, and the length of the video of each measurement was set to 30 s.

Western blot analysis
Protein concentrations of EV samples were determined using the micro-bicinchoninic  Figure   S2 for uncropped Western blots.

Transmission electron microscopy (TEM)
TEM was used to investigate the microstructure of NPC-EVs. Briefly, formvar-coated TEM grids (copper, 150 hexagonal mesh, Science Services, Munich, Germany) were put on the top of a droplet of the respective EV fraction and incubated for 10 min. Then, the grids were washed five times by incubation for 2 min in PBS, followed by similar incubations with ultra-pure water. For contrast, the grids were incubated for 5 min on droplets of uranylacetateoxalate, followed by 5 min incubation on droplets of a 1:9 dilution of 4 % uranylacetate in 2 % methylcellulose. These solutions were prepared as described as before [35]. After draining the methylcellulose from the grids using filter paper and drying of the methylcellulose film as previously described, samples were imaged with a LEO912 transmission electron microscope (Carl Zeiss Microscopy, Oberkochen, Germany) and images were taken using an on-axis 2k CCD camera (TRS-STAR, Stutensee, Germany). Up to two missed tryptic cleavages and methionine oxidation as a variable modification were allowed for. Search tolerances were set to 10 ppm for the precursor mass, 0.05 Da for fragment masses, and ESI-QUAD-TOF specified as the instrument type. For further information about the mass spectrometric analyses, please refer to supplementary material "mass spectrometric analyses" attachment.

NPC-EV RNA isolation and qRT-PCR
In order to investigate whether or not NPC-EVs contain distinct sets of miRNAs that might be responsible for the biological effects of EVs on neuroprotection or neuroregeneration, and reverse 5'-GGGCCATGCTAATCTTCTCTG-3'. The PCR cycling included reverse transcription stage at 42°C for 5 min and 95°C for 3 min followed by amplification stage at 95°C for 10 s and 58°C for 20 s. At the melting curve stage, the temperature was set to 95°C for 5 s followed by 65°C for 1 min and set the acquisition mode to continuous, duration mode set to 5-10 acq/°C, the temperature was 97°C at the end. The cooling stage was set to 40°C for 10 s. miRNA expression was quantified using the 2 −ΔCt method

Cerebral organoids and oxygen-glucose deprivation (OGD assay)
Cerebral organoids were generated from in-house generated iPSCs (hiPS-G1) [42] embedded A drop in the blood flow of more than 80 % to the baseline was considered to indicate successful surgery. Sixty minutes after monofilament insertion, the reperfusion was initiated by monofilament removal, and the LDF recordings were continued for an additional 15 min before the wounds were carefully sutured. The amount of NPC-EVs applied to organoids was calculated by the following process. The EVs from 432x10 6 NPCs were diluted to 500 µl with PBS, and each microlitre contains 8.64x10 5 cell equivalent EVs (8.64x10 5 cell equivalent/µl, 43.2 µg/µl). The mice were exposed to MCAO followed by administration of normal saline

Flow Cytometry
Single-cell suspensions were prepared for flow cytometry. The mice were exposed to MCAO followed by administration of normal saline (control), NPC-EVs low (EVs equivalent to For data analysis, FlowJo v10.0 software was used. Please also refer to Supplementary Table   S2 and Supplementary Figure S1 for further information.

Analysis of post-stroke motor coordination deficits
The mice were trained on days 1 and 2 before the induction of stroke to ensure proper test behavior. The tests for analysis of motor coordination were performed at the time points given using the tight rope test, the balance beam test, and the corner turn test, as previously described [43]. The tight rope test and the balance beam test were performed three times on each test day, and the mean values were calculated. For the balance beam test, the readout parameter was the time until the mice reached the platform, with a maximal testing time of 60 s. Assessment of the tight rope test results was done using a validated score ranging from 0 (minimum) to 20 (maximum). The corner turn test included 10 trials per test day, during which the laterality index (number of left turns / 10) was calculated. The details of the tight rope test score sheet can be found in Supplementary Table S3.
Immunofluorescence staining

Statistical analysis
For comparison of two groups, the two-tailed independent Student's t-test was used.
For comparison of three or more groups, a one-way analysis of variance (ANOVA) followed by the Tukey's post-hoc-test and, if appropriate, a two-way ANOVA were used. G*Power was used to calculate the power of the experiment and GraphPad Prism was used for statistics.
Unless otherwise stated, data are presented as mean with SD values. A p-value of <0.05 was considered statistically significant.

Characterization of NPC-EVs
Since the optimal procedure for EV enrichment remains uncertain [19], we systematically analyzed NPC-EVs using both the PEG method and ultracentrifugation only.
The subsequent characterization of such enriched NPC-EVs included transmission electron microscopy (TEM), nanosight tracking analysis (NTA), mass spectrometry, and Western blotting. Western Blot analysis for selected EV biomarkers revealed that CD63, TSG101, TAPA1, and Alix were present in NPC-EVs obtained from both PEG enrichment and ultracentrifugation only (Fig. 1A-B). No difference was observed with regard to quantitative analysis of these proteins.
NTA revealed the distribution of NPC-EVs to be in the range of 30 nm to 300 nm, which is typical of exosomes and microvesicles alike (Fig. 1C). The concentration between the two isolation methods was similar (18x10 9 in the PEG method compared to 15x10 9 in the ultracentrifugation only method). Along with the NTA, experiments using TEM revealed a typical EV morphology (Fig. 1D). Furthermore, no other types of vesicles were found in our samples, suggesting that NPC-EVs were predominantly exosomes and microvesicles.
By using mass spectrometric analyses, we successfully detected proteins which are crucial for EV biogenesis or are known to be associated with stimulated angiogenesis and neurogenesis such as HSP70, both in PEG-enriched and in ultracentrifugation-enriched NPC-EVs (please also refer to the supplementary section). EV cytosolic proteins, transmembrane proteins, or GPI-anchored proteins, which are crucial for EV biogenesis, such as ANXA6, SDCPB, HSPA8, TSG101, CD81, and CD9 were found in the EV samples ( Fig. 2A-B). There was no significant difference between the two groups with respect to cytosolic or transmembrane proteins. Although calnexin was considered as a negative control for EVs, calnexin can also be found in some subtypes of exosomes [19]. Other subtypes of exosomal markers such as GM130 and CYC1 were close to the detection threshold or not detectable at all (Fig. 2C). Furthermore, strict EV negative markers (as purity control) such as APOA1, ultracentrifugation group (Fig. 2D), suggesting a sufficiently high level of purification in EVs enriched with either PEG or ultracentrifugation.
Since miRNAs support angiogenesis, neurogenesis, and neuroprotection [44], we chose some of those miRNA candidates which have been identified as beneficial in the aforesaid aspects. Indeed, typical miRNA candidates were found in NPC-derived EVs, with the subtypes miR-20a, miR-26b, and miR-124 being at the highest concentrations (Fig. 2E).
Conversely, miR-133, and miR-145 were hardly found in NPC-derived EVs at all. The isolation method, however, did not significantly affect these miRNA levels in NPC-EVs.
Since both enrichment procedures did not significantly differ between each other for the majority of the readout parameters analyzed for NPC-EV characterization, further experiments were performed using the PEG method only. In this context, the PEG approach is more feasible than ultracentrifugation, allowing the handling of large volumes of conditioned medium with ease.

NPC-EVs protect cerebral organoids from oxygen-glucose-deprivation (OGD)
After having characterized the aforementioned EVs, we then established an in vitro model of OGD in cerebral organoids, which better reflect the physiological situation than neuronal monolayer cultures do. We first tested the time course of OGD-induced cell death of cerebral organoids by exposing the organoids to either 8 h or 10 h of OGD, followed by 24 h of reoxygenation under standard cell culture conditions (Fig. 3A-B). Cell death rates under these conditions were significantly increased in organoids exposed to both 8 h and 10 h of OGD when compared to cerebral organoids that were kept under standard culture conditions (Fig. 3B). However, no significant difference was found between the two OGD groups themselves. Accordingly, we chose an OGD exposure of 8 h for the following experiments.
Cerebral organoids were then treated with low, medium or high concentrations of NPC-derived EVs, and the cell death rate was measured after 8 h of OGD followed by 24 h of reoxygenation ( Fig. 3C). Indeed, exposure to all three concentrations of NPC-EVs significantly reduced the cell death rate of cerebral organoids under these conditions, as assessed by TUNEL staining.

Delivery of NPC-EVs reduces post-ischemic motor coordination impairment
In light of the aforementioned in vitro data on cerebral organoids, we tested the hypothesis that NPC-EVs improve neurological recovery after cerebral ischemia in mice.
Following a previously published protocol on MSC-EVs [17], NPC-EVs of different dosages (low, medium, and high) were systemically administered on days 1, 3, and 5 post-stroke, with MSC-EVs serving as internal controls (Fig. 4A-C). Administration of a medium dosage of both NPC-EVs and MSC-EVs resulted in significantly better test performance of these animals in the tight rope test as well as in the corner turn test when compared to controls (Fig. 4B-C). Of note, NPC-EVs were not inferior to MSC-EVs, and the better test performance of mice treated with either NPC-EVs or MSC-EVs was long-lasting and thus stable until the end of the observation period of 84 days. In the balance beam test, however, the beneficial effects of NPC-EVs and MSC-EVs delivered at a medium dose were only transiently effective (Fig. 4A).
Delivery of NPC-EVs or MSC-EVs at low or high dosages only partially induced neurological recovery in these three tests, if at all. The laser Doppler flow was used to ensure the quality of the MCAO model in each group. Each group showed a significant blood flow drop during the surgery, and there was no significant difference between the various treatment groups (Fig.   4D).
Neurological recovery does not necessarily imply an effect on brain tissue injury or brain regeneration and vice versa. We subsequently analyzed neuronal survival in the ischemic striatum at 84 days after the stroke. In line with the reduction of neurological impairment, increased neuronal densities were found in mice treated with medium doses of both NPC-EVs or MSC-EVs (Fig. 5A), again showing no difference between these two groups.
Low and high doses of NPC-EVs or MSC-EVs were not effective. Conclusively, NPC-derived EVs reduce post-stroke brain injury on both the histological and the functional level and are not inferior to MSC-derived EVs. It is important to note that the therapeutic effect in EV formulation was highly dose dependent.

NPC-EV delivery stimulates post-stroke neuroregeneration and axonal plasticity
As stated before, neurological recovery is not always associated with histological changes, and the mechanisms that lead to enhanced neurological recovery are diverse. In this context, we hypothesized that NPC delivery might stimulate endogenous repair mechanisms of the brain, including increased levels of neurogenesis. Taking into account that neurogenesis also takes place in the adult mammalian brain, with endogenous stem cells being stimulated upon induction of cerebral ischemia, the application of NPC-EVs might positively interfere with this process. Indeed, analysis of the cell proliferation marker BrdU showed significantly increased levels of BrdU + cells (Fig. 5B) in the NPC-EV medium group.
The co-expression analysis with the proliferation marker BrdU and the neuronal marker NeuN revealed increased levels of NeuN + /BrdU + cells on day 84 within the ischemic striatum of animals treated with a medium dosage of NPC-EVs (Fig. 5C). On the contrary, the relative amount of BrdU + cells expressing the immature neuronal marker Dcx was not affected by NPC-EV-treatment (Fig. 5D).
Neuroregeneration is a complex process that is not only limited to neurogenesis. We therefore investigated the extent of axonal plasticity on day 84 in the post-ischemic brain (Fig.   6), using contra-lateral stereotactic injections of BDA in the non-impaired cortex. Again, delivery of a medium dose of NPC-EVs but not low or high doses of NPC-EVs significantly enhanced axonal plasticity in these mice when compared to the control group (Fig. 6).
Treatment with a medium dosage of MSC-EVs yielded similar effects with regard to neuroregeneration and axonal plasticity as observed for NPC-EVs, suggesting again that NPC-EVs are not inferior to MSC-EVs.

NPC-EVs reverse peripheral post-stroke immunosuppression
The pathophysiology of cerebral ischemia comprises of a complex string of diverse inflammatory signaling cascades, not solely being harmful to the surrounding ischemic tissue [45]. In our previous study, MSC-EVs did not affect the immune response in the central nervous system but reversed the post-ischemic immunosuppression in the peripheral blood seven days after stroke. In NPC-EVs, we saw similar effects. Consequently, NPC-EV treatment of mice with either dosage did not affect leukocytes (CD45 high ), monocytes (CD45 high CD3 -CD11b + ), B cells (CD45 high CD3 -CD19 + ) or T cells (CD45 high CD3 + ) within the ischemic CNS (Fig. 7A-D). Flow cytometry analysis in the blood of mice treated with a medium but not a low or high dosage of NPC-EVs revealed significantly increased levels of both B lymphocytes and T lymphocytes when compared to the control group (Fig. 7E-H).

NPC-EVs predominantly distribute in peripheral organs
The majority of MSCs and other transplanted cells do not reach the brain, but are trapped in extracranial organs [46]. Even though EVs are known to pass the blood-brain barrier [47], the fact that NPC-EVs predominantly modulated the peripheral but not the central immune system (Fig. 8) might suggest that the majority of EVs do not reach the brain, either.
Since different administration methods and conditions might affect the biodistribution patterns of NPC-EVs, we compared two different delivery routes, i.e., femoral vein injection and retroorbital injection under both ischemic and non-ischemic conditions. The biodistribution of NPC-EVs was similar in different methods and different conditions (Fig. 8). NPC-EVs were not only found in peripheral organs such as the liver and the lung but also found in the brain.
However, most of these EVs were detected in the liver and in the lung when compared to the brain. There was no difference between the liver and the lung with regard to NPC-EV biodistribution patterns.

Discussion
Although EVs have recently been recognized as potential therapeutic tools in the treatment of stroke [48], previous work has almost exclusively focused on the application of MSC-derived EVs only [17]. As such, the relevance of the stem cell source for EV enrichment remains elusive, although recent data on pluripotent stem cell-derived NSC-EVs has become available [49,50]. The present study elucidated whether or not EVs have a cell-type independent therapeutic potential against stroke that is not restricted to MSC-derived EVs. In light of various EV enrichment procedures, i.e., ultracentrifugation, precipitation, chromatography, or density gradient separation [52], the optimal enrichment technique is still a matter of debate [19]. The pros and cons of each technique have to be thoroughly balanced when being applied. In that respect, precipitation methods using PEG or others offer quick and easy handling of large cell supernatant volumes for EV enrichment, although none of these techniques currently qualify for EV enrichment under GMP standards. Previous work from our own group systematically analyzed the PEG precipitation approach in direct comparison to standard EV enrichment procedures on HEK293T cells [33]. The latter revealed the PEG approach to be not inferior to standard EV enrichment procedures using HEK293T cells.
Compared to the ultracentrifugation method, the PEG method can concentrate a high volume of conditioned medium in low centrifugation force (4,500 g) which can reduce EV damage from shear force during the ultracentrifugation process. Although basic EV properties such as the size should be unaffected when dealing with different cell sources [53], the optimal concentrations of these cell sources under different isolation methods remains a matter of debate. To the best of our knowledge, PEG precipitation has not been used for the enrichment of SVZ-derived NPC-EVs. In order to exclude an impact of PEG precipitation on NPC-derived EVs, a detailed characterization of the latter was therefore performed.
Using PEG 6000, we successfully isolated EVs from NPC conditioned medium. The purification rates obtained by this method were high for NPC-EVs, as indicated in the mass spectrometric analysis result. Especially, EV negative markers such as APOA1, UMOD, and ALB were close to the detection threshold or not detectable at all in our PEG method group and ultracentrifugation only group. Likewise, the distribution patterns were similar to EVs enriched with ultracentrifugation only, which is still regarded as a gold standard [19]. Our observations are in line with previous work from our group on the comparison of the PEG method and direct ultracentrifugation or differential centrifugation, indicating that PEG does not significantly affect the purity of EVs [33], although artifacts and aggregations still occur as observed in the TEM analysis. As a matter of fact, previous studies on EV application under pathological conditions such as Alzheimer's disease [54] and stroke [55], have either focused on exosomes only or used the general term of EVs, which might be more convenient in light of a therapeutic approach. The mass spectrometric analyses on NPC-EVs as performed in the present work indicated expression patterns of both exosomal markers and microvesicle markers [19]. Some of these proteins found in our EV samples are well-known key mediators of neuroprotection and neurogenesis. The heat shock protein HSP70, for instance, was highly abundant in NPC-derived EVs. HSP70 has frequently been described to mediate a plethora of signaling cascades, all of which contribute to an enhanced resistance of neural cells under hostile hypoxic or ischemic conditions [56]. Under such circumstances, HSP70 has been shown to modify oxidative stress and proteasomal activity of cerebral tissue exposed to hypoxic or ischemic injury, resulting in enhanced cell survival and increased neurological recovery of stroke rodents [56]. The increased resistance of cerebral organoids exposed to OGD injury treated with NPC-EVs might, therefore, be at least partially mediated by proteins such as HSP70 and others.
Since pluripotent stem cell-derived NSC-EVs have recently been proven to increase tissue rescue and functional outcomes in translational murine and porcine stroke models [49,50], our data support the hypothesis that EVs from different cell sources including SVZ-derived NPCs could induce neurological recovery as well. As indicated before, EVs do not only carry diverse sets of proteins but also contain DNA and non-coding RNA, among which are miRNAs of particular interest. The latter has been extensively studied under physiological and pathological conditions alike, not only related to cerebral ischemia where they serve both as biomarkers and therapeutic tools [57]. Screening for selected miRNAs to be likely expressed in NPC-derived EVs revealed enhanced levels of miR-124, which is among the most abundant miRNAs in the adult mammalian brain [58,59]. MiR-124 affects a plethora of signaling molecules such as the recently identified inhibition of deubiquitination of Usp14, significantly contributing to the reduction of post-stroke brain injury in rodents [60]. Increased levels of miR-124 and others such as the microRNA 17-92 cluster found in NPC-EVs, known to contribute to neuroprotection or enhanced neuroregeneration [61][62][63][64][65][66][67], might thus further enhance the resistance of cells and tissues against hypoxia or ischemia. Since EVs are able to transfer cargo like miRNAs to target cells [68], it is fair enough to hypothesize that NPC-EVs yield protection of cerebral organoids and brain tissue using this way of action. In this context, recent research has demonstrated that oligodendrocytes are able to transfer mitochondria toward neurons via microvesicles upon induction of hypoxia, a mechanism that might also be of relevance under the conditions chosen for the present work [69][70][71]. Elucidating such a precise mechanism was, however, beyond the scope of the present work which set emphasis on the therapeutic potential of NPC-EVs rather than on mechanisms involved in such a process.
The therapeutic potential of stem cells and the different mechanisms being involved greatly depend on delivery routes and transplantation timing [3]. Indeed, systemic delivery of MSCs or NPCs under experimental stroke conditions reduces post-stroke brain injury, enhances neurological recovery, and stimulates neuroregenerative processes [3,72,73].
Systemic delivery of MSC-derived EVs, likewise, mediates the aforementioned effects in a similar fashion [17]. Both timing and the delivery route were chosen following a previously established protocol for the intravenous application of MSC-derived EVs [17], although future studies might enhance the therapeutic time frame for EVs even further. Contrary to grafted stem cells, which predominantly do not reach the brain and which up in peripheral organs such as the lung and the liver rather than in the brain. However, additional studies regarding homing and targeting of ectopic EVs are required to elucidate these underlying mechanisms further.
In our study, medium dosages of NPC-EVs (and also MSC-EVs) showed optimal outcomes when compared to either low or high dosages of EVs.

Disclosure/conflict of interest
The authors declare to have no conflict of interest of any kind.

Ethical approval:
This article does not contain any studies with human participants performed by any of the     condition), i.e., NPC-EVs low (EVs equivalent to 2x10 5 NPCs), NPC-EVs medium (EVs equivalent to 2x10 6 NPCs), and NPC-EVs high (EVs equivalent to 2x10 7 NPCs). Control organoids were exposed to OGD only without EV treatment, whereas "normal" refers to cerebral organoids kept under standard cell culture conditions. Scale bars: 20 µm.  Motor coordination was evaluated using the balance beam test (A), the tight rope test (B), and the corner turn test (C) at 1, 7, 14, 28, 56, and 84 days after cerebral ischemia. All animals were accordingly trained before the induction of stroke in order to ensure proper test performance, i.e., test results before induction of stroke are given as pre-stroke data. Mice