Extracellular Vesicles from Vascular Endothelial Cells Promote Survival, Proliferation and Motility of Oligodendrocyte Precursor Cells

We previously examined the effect of brain microvascular endothelial cell (MVEC) transplantation on rat white matter infarction, and found that MVEC transplantation promoted remyelination of demyelinated axons in the infarct region and reduced apoptotic death of oligodendrocyte precursor cells (OPCs). We also found that the conditioned medium (CM) from cultured MVECs inhibited apoptosis of cultured OPCs. In this study, we examined contribution of extracellular vesicles (EVs) contained in the CM to its inhibitory effect on OPC apoptosis. Removal of EVs from the CM by ultracentrifugation reduced its inhibitory effect on OPC apoptosis. To confirm whether EVs derived from MVECs are taken up by cultured OPCs, we labeled EVs with PKH67, a fluorescent dye, and added them to OPC cultures. Many vesicular structures labeled with PKH67 were found within OPCs immediately after their addition. Next we examined the effect of MVEC-derived EVs on OPC behaviors. After 2 days in culture with EVs, there was significantly less pyknotic and more BrdU-positive OPCs when compared to control. We also examined the effect of EVs on motility of OPCs. OPCs migrated longer in the presence of EVs when compared to control. To examine whether these effects on cultured OPCs are shared by EVs from endothelial cells, we prepared EVs from conditioned media of several types of endothelial cells, and tested their effects on cultured OPCs. EVs from all types of endothelial cells we examined reduced apoptosis of OPCs and promoted their motility. Identification of the molecules contained in EVs from endothelial cells may prove helpful for establishment of effective therapies for demyelinating diseases.


Introduction
Demyelination as a result of white matter ischemia causes loss of brain functions [1,2], but there is no specific treatment so far. We previously showed that transplantation of brain microvascular endothelial cells (MVECs) greatly stimulated remyelination in the white matter infarct culture dish, supplemented with forskolin (Sigma) and CNTF (PeproTech). OPC proliferation was induced by the addition of NT-3 (PeproTech) and PDGF-AA (PeproTech) as mitogens. Cells were grown in 5% CO 2 at 37°C. After several days, OPCs were harvested from dish using TrypLE Express (Gibco), washed, and suspended in the culture medium. OPCs were plated on PDL-coated 8-well slide glasses containing proliferation medium at the density of 12000 cells/ cm 2 .
Primary culture of rat brain microvascular endothelial cells. Brain microvascular endothelial cells (MVECs) were prepared from adult rat cerebra [3,12]. In brief, the meninges and superficial blood vessels were removed, and the cortical tissue was digested with 0.15% dispase (Gibco) solution, triturated, and centrifuged. The pellet was digested again with 1 mg/ml of collagenase/dispase (Roche), washed, loaded onto continuous density gradient of 30% Percoll (GE Healthcare), and centrifuged. The fraction containing the microvessels was collected and centrifuged. The pellet was resuspended in the culture medium (EGM-2, Lonza) containing 4 μg/ ml puromycin (Sigma) [13], and the cells were plated into collagen type I-coated dishes (Iwaki, Japan). Cultures were maintained in 5% CO 2 at 37°C. After treatment with puromycin for 48 h, it was removed from the culture medium.

Analysis of the effect of MVEC-CM on cultured OPCs
To prepare MVEC-CM, MVECs (90-95% confluence) were washed with PBS, and then grown in serum-free medium for 24 h. CM was collected, and centrifuged to remove cell debris. The supernatant was filtered using 0.45μm filter before use in OPC experiments. To deplete EVs from MVEC-CM, MVEC-CM was ultracentrifuged at 130,000 × g at 4°C overnight, and the supernatant was collected as EV-depleted MVEC-CM (EV-dep-MVEC-CM). Neat CM was used in this experiment. OPCs plated on 8-well glass slides were wash with PBS, and then cultured in MVEC-CM or EV-dep-MVEC-CM for 3 days. As control, OPCs were cultured in serum-free medium.

Isolation of EVs from culture media
For EV isolation, exosome-depleted FBS (System Biosciences) was used for cultures of ECs and Rat-1 cells to prevent contamination by EVs derived from FBS. After washing with PBS, cells were cultured in medium containing exosome-depleted FBS (MVECs, HAECs and HUVECs: EGM-2 containing 2% exosome-depleted FBS, HMVECs: EGM-2MV containing 5% exosomedepleted FBS, bEnd.3 and Rat-1: DMEM containing 10% exosome-depleted FBS) for 72 h. EV precipitation from conditioned media was performed using the exosome precipitation solution, ExoQuick-TC, (System Biosciences) according to the manufacturer's protocol. Conditioned media were centrifuged at 3,000 × g for 15 minutes to remove cells and debris. The supernatant was mixed with ExoQuick solution, and refrigerated overnight. Samples were centrifuged at 1,500 × g for 30 minutes, and supernatant was removed. Samples were centrifuged again at 1,500 × g for 5 minutes, and all traces of supernatant were removed. EV pellet was resuspended in a small volume of buffer or serum-free medium depending on the following analyses.

Analysis of the effect of EVs on cultured OPCs
Isolated EVs were suspended in serum-free medium at the protein concentrations of 3, 12.5, 50, and 200 μg/ml, and added to OPC cultures. OPCs plated on 8-well glass slides were wash with PBS, and then cultured in serum-free medium containing EVs for 2 days. As control, OPCs were cultured in serum-free medium.

Cell survival assay
Cell death with the typical morphological features of apoptosis including pyknotic nuclei was assessed by staining cell nuclei with Hoechst 33342 (Sigma, 2.5 μM) for 40 min in 5% CO 2 at 37°C. After incubation, samples were observed using a fluorescent microscope (Axioplan2, Zeiss) with Cooled CCD Camera (DP73, Olympus), and the number of pyknotic cells in each sample was counted.

BrdU incorporation assay
To assess cell proliferation, bromodeoxyuridine (BrdU; 10 μM) (Roche Diagnostics) was added to the cultures for the last 4 h of culture, followed by fixation and staining with rat monoclonal BrdU antibody (abcam; 1:1000). Propidium iodide (Sigma, 2 μg/ml) was used for nuclear staining. Samples were observed using a fluorescent microscope (Axioplan2, Zeiss) with Cooled CCD Camera (DP73, Olympus), and the number of BrdU-positive cells in each sample was counted.
Cell motility assay using reaggregated OPCs OPC migration was measured using reaggregated OPCs [14]. OPC reaggregates were formed by plating OPCs on Petri dish (Corning) with shaking for 6 h. The reaggregated OPCs were then plated onto PDL-coated 8-well glass slides in MVEC-CM, EV-dep-MVEC-CM and serum-free medium containing EVs with 10 μM AraC. After incubation at 37°C in 5% CO 2 for 16 h, the images of OPCs were captured using an inverted microscope (Primo-Vert, Zeiss). The distance of migration was calculated by measuring the change in size of the reaggregates over time, subtracting the average initial size of the reaggregates, and dividing the remaining distance in half. More than three aggregates in each condition were measured in three independent experiments.

Protein assay and Western blot analysis
Protein concentration of solution including isolated EVs was measured using Protein Quantification Assay Kit (Takara Bio) according to the manufacturer's protocol.

Nanoparticle tracking analysis (NTA)
Isolated EVs were analyzed by nanoparticle tracking using NanoSight NS300 system (Malvern Instruments Ltd, Marvern, UK). EV samples were measured five times at 24 to 27°C, and particle sizes and concentrations in the samples were estimated.

Statistical analysis
Statistical analysis was performed using R software (ver 3.2.2). Differences between groups were assessed with one-way analyses of variance followed by the post hoc Tukey-Kramer test. Student's t-test was used to evaluate the differences between two groups. The error bars represent the standard errors. Cell counting was performed by persons unaware of the profile of samples.

Depletion of EVs from MVEC-CM reduced its effect on OPC survival and motility
To examine whether EVs contained in MVEC-CM contribute to its inhibitory effect on apoptosis of cultured OPCs, OPCs were cultured in EV-depleted MVEC-CM (EV-dep-MVEC-CM).
In MVEC-CM, many OPCs survived for 3 days with typical morphological features of healthy cells ( Fig 1A). In contrast, in serum-free medium (control) and EV-dep-MVEC-CM, many OPCs died with the typical morphological features of apoptosis including pyknotic nuclei. As we reported previously, MVEC-CM had no apparent effect on OPC differentiation [4].
To analyze cell survival, OPCs were stained with Hoechst 33342, a cell-permeant nuclear staining dye, and the nuclear morphology was observed. The proportion of pyknotic (apoptotic) nuclei (red arrowheads) in MVEC-CM and EV-dep-MVEC-CM were both smaller when compared to control (Fig 1B and 1E). The proportion of pyknotic nuclei in EV-dep-MVEC-CM, however, were significantly larger when compared to MVEC-CM ( Fig 1E).
We then examined the effect of MVEC-CM on OPC proliferation. The proportion of BrdUpositive cells in MVEC-CM was the same as that in control, but that in EV-dep-MVEC-CM was significantly smaller when compared to MVEC-CM and control (Fig 1C and 1F).
We also examined the effect of MVEC-CM on OPC motility. MVEC-CM greatly promoted motility of OPCs when compared to control and EV-dep-MVEC-CM, though EV-dep-MVEC-CM still showed a small but significant promoting effect on OPC motility (Fig 1D and 1G).
Taken together, MVEC-CM promoted OPC survival and motility, and removal of EVs from MVEC-CM by ultracentrifugation significantly reduced its effect on OPC survival and motility.
EVs from MVEC-CM were taken up by cultured OPCs and promoted their survival, proliferation, and motility As the above experiments suggested a possibility that MVEC-derived EVs influence the behavior of OPCs, we isolated EVs from MVEC-CM to examine directly whether EVs affect cultured OPCs. When EVs prepared from MVEC-CM (MVEC-EVs) were analyzed by Western immunoblotting for CD63 and CD9, transpanin family proteins, they were positive for these exosome markers (Fig 2A).
When MVEC-EVs were labeled with PKH67, a fluorescent dye, and added to OPC cultures, PKH67-labeled EVs were immediately taken up by OPCs, and many small vesicles were found within OPCs (Fig 2B). This result supports the possibility that MVEC-EVs affect cultured OPCs.
We then examined the effect of MVEC-EVs on cultured OPCs. After 2 days in culture in the presence of EVs (50 μg/ml protein), there was a significant difference in the morphology of OPCs. In the presence of MVEC-EVs, OPCs looked much healthier with sturdy processes when compared to control (Fig 3A). To estimate cell death, OPCs were stained with Hoechst 33342 dye. There were significantly less pyknotic nuclei in the presence of MVEC-EVs when compared to control (Fig 3B and 3E), suggesting that MVEC-EVs inhibit apoptosis of OPCs.
To determine the proportion of proliferating cells, BrdU incorporation assay was used. More BrdU-positive OPCs were seen in the presence of MVEC-EVs when compared to control ( Fig  3C and 3F), suggesting MVEC-EVs promote OPC proliferation. We also examined the effect of MVEC-EVs on OPC motility and found that MVEC-EVs greatly promoted the motility ( Fig  3D and 3G). Taken together, MVEC-EVs promoted OPC survival, proliferation, and motility.
Next, we examined the dose-dependent effects of MVEC-EVs on OPC behavior. Isolated MVEC-EVs were suspended in serum-free medium and add to OPC culture at 0, 3, 12.5, 50, and 200 μg/ml proteins. After 2 days in culture, the proportion of pyknotic nuclei was reduced at doses 50 μg/ml (Fig 4A). The proportion of BrdU-positive OPCs at doses 50 μg/ml increased as compared with lower doses (0, 3, 12.5 μg/ml) (Fig 4B). For OPC motility assay, reaggregated OPCs were cultured overnight at 0, 3, 12.5, 50, and 200 μg/ml proteins. MVE-C-EVs increased the distance of OPC migration in a dose-dependent manner (Fig 4C). These results clearly revealed that MVEC-EVs promoted OPC survival, proliferation and motility in a dose-dependent manner.

EVs derived from endothelial cells promoted OPC survival, proliferation and motility
We then examined whether these effects on cultured OPCs are shared by EVs from endothelial cells. EVs were prepared from various types of endothelial cells (bEnd.3: immortalized mouse . EVs were also prepared from rat fibroblast-like cell line (Rat-1). EVs prepared from CM of cultures of each cell type were suspended in serum-free medium and added to OPC culture at 50 μg/ml of EV proteins, and maintained for 2 days. As control, OPCs were cultured in serum-free medium without EVs.
After 2 days in culture, in the presence of HAEC-and HUVEC-EVs, the shape of OPCs were similar to that in MVEC-EVs (Fig 3A), having sturdy processes, as shown in Fig 5A. In the presence of MVEC-EVs, HAEC-EVs, HMVEC-EVs and HUVEC-EVs, pyknotic nuclei were fewer when compared to bEnd.3-EVs and Rat-1-EVs, although the latter two EVs also significantly reduced pyknotic nuclei when compared to control (Fig 5B and 5E). Treatment with EC-derived EVs, except for bEnd.3-EVs, significantly inhibited apoptotic cell death of OPCs when compared to Rat-1-EVs. In the presence of EC-derived EVs, except for HMVE-C-EVs, the proportion of BrdU-positive cells was significantly larger when compared to control (Fig 5C and 5F). There was no significant difference in the number of BrdU-positive cells in the presence of HMVEC-EVs or Rat-1-EVs when compared to control (Fig 5F). Treatment of reaggregated OPCs with EC-derived EVs promoted their motility, and OPCs spread radially from initial reaggregates, while Rat-1-EVs had no effect on OPC motility (Fig 5D and 5G).

Discussion
We confirmed our previous study reporting that MVEC-CM reduced apoptotic cell death of OPCs [4], and found that MVEC-CM promoted their motility (Fig 1). ECs are active and dynamic cells producing various biologically active molecules in response to various stimuli. It is well known that FGF-2, PDGF and VEGF are among important soluble proteins playing roles in OPC survival, proliferation and differentiation [15][16][17][18][19][20][21], and are released from ECs [21][22][23][24][25][26]. In addition, adhesion proteins such as laminin and fibronectin also affect OPC behavior [27,28], and are secreted by ECs [29]. MVEC-CM possibly contains these molecules, and these molecules in soluble form may exert various effects on cultured OPCs. In this study, we examined the effect of EVs prepared from rat brain MVECs on the behavior of rat OPCs, and obtained several lines of evidence supporting a role for MVEC-derived EVs in the cell-cell communication between MVECs and OPCs. First, inhibitory effect of MVEC-CM on OPC apoptosis were significantly reduced by removal of MVEC-EVs by ultracentrifugation. Second, fluorescently labeled MVEC-EVs are readily taken up by cultured OPCs. And third, MVE-C-EVs promoted survival, proliferation and motility of cultured OPCs.
MVEC-CM showed no enhancement of OPC proliferation, while MVEC-EVs promoted their proliferation. There are two possibilities for this discrepancy. One is that we used neat CM for our experiments and the concentrations of EVs contained in CM were too low to exert their effect. The other is that some soluble factors inhibitory for OPC proliferation are contained in CM and these negate the promoting effect of EVs. Although the finding that EV-dep-MVEC-CM reduced OPC proliferation when compared to control supports the latter possibility, further studies are necessary to test these possibilities.  Recently there are several reports on EVs in the central nervous system (CNS). It has recently been reviewed that neuronal EVs play roles in neuronal-glial communication and contribute to neuronal development and disease mechanisms [30,31]. The in silico interactomics analyses, which were based on molecular profile of EVs derived from human brain microvascular endothelial cells (HBECs), showed that HBEC-EVs may engage in numerous cell-surface interactions with astrocytes and neurons [32]. Thus there is a possibility that similar EV-mediated communication might occur between ECs and OL-lineage cells.
EVs released from ECs are reported to contain various miRNAs [33,34]. It has already been reported that some miRNAs are closely related to the development of OL-lineage cells [35][36][37][38]. MicroRNA (miR-17-92) cluster is important in cell cycle, proliferation, apoptosis and other pivotal processes [39]. It is known that this microRNA cluster is highly expressed in ECs [40], and that, in OPCs, this cluster promotes cell survival and proliferation by influencing Akt signaling [41]. Therefore, there is a possibility that members of miR-17-92 cluster contained in MVEC-EVs promote survival and proliferation of OPCs. MicroRNA-221/222 (miR-221/222) are also highly expressed in ECs, and one of their target genes is the cyclin-dependent kinase inhibitor, p57/Kip2 (Cdkn1c) [42]. In OL-lineage cells, p57/Kip2 plays a crucial role in cell proliferation by inhibiting CyclinE-cdk2 complex formation. Reducing p57/Kip2 in OPCs, therefore, promotes their proliferation [43]. So there is another possibility that miR-221/222 contained in MVEC-EVs reduces expression of p57/Kip2 in OPCs and thereby promotes OPC proliferation. Mass spectrometric analysis revealed that various types of proteins are present in EVs released from ECs [32,33]. Proteomics analyses showed that cell surface proteins, including adhesion molecules, extracellular matrix (ECM) molecules and other cell-cell interacting molecules, were present in EVs [32,44]. It is well known that ECM molecules regulate various aspects of cellular behavior [45], and it was reported that ECM molecules affected the growth properties of OPCs [28]. Thus the effect of MVEC-CM on OPCs observed in this study may be mediated by ECM molecules present in MVEC-EVs.
Several cell types, derived from different species and organs, were used in this study. EVs prepared from all cell types we examined reduced apoptosis of cultured OPCs (Fig 5). EVs secreted by primary cultured ECs (MVEC, HAEC, HMVEC, HUVEC) were more effective on OPC survival than those from immortalized ECs (bEnd.3) or fibroblast-like cells (Rat-1). This result suggests the possibility that EVs prepared from each cell type contain the same molecules with a pro-survival activity on OPCs. Differences in effects of EVs on OPC survival may reflect the quantity of the molecules present in EVs, and EVs secreted by primary cultured ECs may contain them in larger quantity. Regarding OPC proliferation, HMVEC-EVs and Rat-1-EVs showed no effect. Rat-1-EVs didn't affect OPC motility either. These results suggest that several different molecules (proteins and/or RNAs) in the EVs were independently or synergistically affect OPC behavior (survival, proliferation and motility). Further studies are necessary to identify the molecules responsible for the effects of MVEC-EVs on OPCs.
Recently therapeutic potentials of EVs for CNS diseases have been suggested [46,47]. Systemic treatment of stroke with EVs derived from multipotent mesenchymal stromal cells (MSCs) promoted neurovascular plasticity and functional recovery after stroke in rats [48]. miR-219, microRNA contained in EVs derived from dendritic cells and serum, stimulated remyelination in rats [49,50]. Our study demonstrates that EVs prepared from ECs promote survival, proliferation, and motility of OPCs. Although it remains undetermined whether EVs released from transplanted MVECs are involved in its beneficial effect on white matter infarct, identification of molecules present in EVs from ECs may prove useful in establishment of effective therapeutic strategy against demyelinating diseases. In this respect, our finding that HUVEC-EVs affect OPC behaviors is promising, as it is relatively easy to prepare EVs from HUVECs in quantity necessary for proteomic analysis and RNA profiling.