M2 microglia-derived extracellular vesicles promote white matter repair and functional recovery via miR-23a-5p after cerebral ischemia in mice

Rationale: White matter repair is critical for the cognitive and neurological functional recovery after ischemic stroke. M2 microglia are well-documented to enhance remyelination and their extracellular vesicles (EVs) mediate cellular function after brain injury. However, whether M2 microglia-derived EVs could promote white matter repair after cerebral ischemia and its underlying mechanism are largely unknown. Methods: EVs were isolated from IL-4 treated microglia (M2-EVs) and untreated microglia (M0-EVs). Adult ICR mice subjected to 90-minute transient middle cerebral artery occlusion received intravenous EVs treatment for seven consecutive days. Brain atrophy volume, neurobehavioral tests were examined within 28 days following ischemia. Immunohistochemistry, myelin transmission electron microscope and compound action potential measurement were performed to assess white matter structural remodeling, functional repair and oligodendrogenesis. The effects of M2-EVs on oligodendrocyte precursor cells (OPCs) were also examined in vitro. EVs' miRNA sequencing, specific miR-23a-5p knockdown in M2-EVs and luciferase reporter assay were used to explore the underlying mechanism. Results: M2-EVs reduced brain atrophy volume, promoted functional recovery, oligodendrogenesis and white matter repair in vivo, increased OPC proliferation, survival and differentiation in vitro. miR-23a-5p was enriched in M2-EVs and could promote OPC proliferation, survival and maturation, while knocking down miR-23a-5p in M2-EVs reversed the beneficial effects of M2-EVs both in vitro and in vivo. Luciferase reporter assay showed that miR-23a-5p directly targeted Olig3. Conclusion: Our results demonstrated that M2 microglia could communicate to OPCs through M2-EVs and promote white matter repair via miR-23a-5p possibly by directly targeting Olig3 after ischemic stroke, suggesting M2-EVs is a novel and promising therapeutic strategy for white matter repair in stroke and demyelinating disease.

time was 5 minutes. The test trail was repeated three times with an interval more than 15 minutes, the time mice stay on the rod (latency to fall) were recorded and the average time was used for data analysis.

Hanging wire test
Hanging wire test was performed as previously described [2,4]. Briefly, the mice were gently placed on a steel wire (55 cm long and 2 mm thick) for adaption, then placed in the middle of the wire with only two fore limbs and they would manage to climb to one side of the wire. The times mice reached one side or fell off the wire were recorded. Once the mice arrived or fell, they would be placed back to the middle of the wire immediately until the 3 minutes test time finished. The total score was calculated by adding the times of arrivals and subtracting the times of falls from the initial score of 10.

Corner test
The mice were subjected to 10 trails of corner test, during which the left or right turns were recorded [5]. Briefly, the mouse was placed to a 30° angle corner formed by two boards in a quiet and dark room, it would turn back when entering into the corner. Only when the mouse moved forward, upward and put one of the fore limbs onto the board then turn back was considered as a trail.
Normal mice would turn to either side equally, while the ischemic mice would tend to turn to the lesion side, which is the left side in our study.

T-maze
T-maze was performed to exam the working memory function of the mice [6]. At 14 or 28 days after tMCAO, the mouse was placed in the start area and free to choose a goal arm. After the mouse entering the chosen arm, quickly and quietly closed the door and confined the mouse in the chosen arm for 15 seconds, then took it out to a cage, rubbed the maze with 75% alcohol, and replaced it in the start area to allow the mouse choose between the familiar and novel arms, if the mouse chose the novel arms it was considered as a correct spontaneous alternation. The alternations were recorded in 10 trails for each mouse.

Step-through passive avoidance test
Step-through passive avoidance test was performed as previously described [7]. Briefly, one day before the examine time point, the mouse was placed into the light zone of the smart cage (AfaSci Research Laboratories, Redwood City, CA) with free access to the dark zone. When mouse entered into the dark zone and stayed more than 2 seconds, it received an electric foot shock. After 24 hours, the mouse was replaced into the light zone without electric stimulation and recorded the moving trace for 5 minutes. The dark zone entry times and dark zone stay time were analyzed automatically using the CageScore 2.6 software.

Supplemental Figures
Flow cytometry showing that that 89.1% of IL-4 treated microglia were Arg-1+ cells and less than 3.6% were CD86+ cells.  Representative HE images showed no detectable morphological changes and toxic effect in the heart, liver, spleen, lung and kidney 21 days after 7 consecutive days of tail injection of microglia-derived EVs in tMCAO mice.