Endothelial cells regulate astrocyte to neural progenitor cell trans-differentiation in a mouse model of stroke

Abstract

The concept of the neurovascular unit emphasizes the importance of cell-cell signaling between neural, glial, and vascular compartments. In neurogenesis, for example, brain endothelial cells play a key role by supplying trophic support to neural progenitors. Here, we describe a surprising phenomenon where brain endothelial cells may release trans-differentiation signals that convert astrocytes into neural progenitor cells in male mice after stroke. After oxygen-glucose deprivation, brain endothelial cells release microvesicles containing pro-neural factor Ascl1 that enter into astrocytes to induce their trans-differentiation into neural progenitors. In mouse models of focal cerebral ischemia, Ascl1 is upregulated in endothelium prior to astrocytic conversion into neural progenitor cells. Injecting brain endothelial-derived microvesicles amplifies the process of astrocyte trans-differentiation. Endothelial-specific overexpression of Ascl1 increases the local conversion of astrocytes into neural progenitors and improves behavioral recovery. Our findings describe an unexpected vascular-regulated mechanism of neuroplasticity that may open up therapeutic opportunities for improving outcomes after stroke.

Fig. 1. Oxygen-glucose deprivation-stimulated brain endothelial cells convert astrocytes into neural progenitor cells.

a Morphologic changes of same astrocytes (AC) at day 0, day 1, day 2 and day 3 post cultured in conditioned neuron media collected from normal neurons (Neuron-con) or OGD-stimulated neurons (Neuron-ogd). b Morphologic changes of same AC at day 0, day 1, day 2 and day 3 post co-cultured with normal pericytes (PCcon) or OGD-stimulated pericytes (PCogd). c Morphologic changes of same AC at day 0, day 1, day 2 and day 3 post co-cultured with normal brain endothelial cells (ECcon) or OGD-stimulated brain endothelial cells (ECogd). d Representative immunostaining of DCX, PSA-NCAM and GFAP in AC or at day 3 post AC co-cultured with ECcon or ECogd. e mRNA levels of Nestin, DCX, TUJ1 and Glast at day 1 post AC co-cultured with ECcon or ECogd; n = 3 biologically independent samples; p < 0.0001. f mRNA levels of Nestin, DCX, TUJ1 and Glast at day 3 post AC co-cultured with ECcon or Ecogd; n = 3 biologically independent samples; p < 0.0001. g mRNA levels of Notch signaling-related genes at day 1 post AC co-cultured with ECcon or ECogd; n = 3 biologically independent samples. h mRNA levels of Notch signaling-related genes at day 3 post AC co-cultured with ECcon or ECogd; n = 3 biologically independent samples. p = 0.0077 for Notch1; p = 0.0017 for Dll1; p = 0.0007 for RBP-Jκ. Experiment was repeated independently 3 times with similar results (a–d). Data are shown as mean ± SEM. Two-way ANOVA with post-hoc Bonferroni adjustment. Scale bar, 50 μm. Source data are provided as a Source Data file.

Fig. 2. Microvesicles derived from OGD-stimulated brain endothelial cells reprogram astrocytes into neural progenitor cells.

a Nanoparticle tracking analysis of microvesicles (MVs) derived from normal brain endothelial cells or OGD-stimulated brain endothelial cells; n = 3 biologically independent samples; p = 0.0354. b mRNA levels of Rab27a in normal brain endothelial cells or OGD-stimulated brain endothelial cells; n = 4 biologically independent samples; p = 0.0321. c Confocal microscopy revealed distribution of Cy3 labeled brain endothelial cells-derived MVs (EC-MVs) in AC at different time points. d Immunostaining images showing a red-to-green color switch of AC that had taken up Cre⁺ MVs derived from brain endothelial cells. e Representative immunostaining of DCX and GFAP in AC treated with vehicle or MVs derived from normal brain endothelial cells (ECcon-MVs) or OGD-stimulated brain endothelial cells (ECogd-MVs). f Quantification of DCX+ cells; n = 6 samples obtained from three independent experiments for vehicle group, n = 8 samples obtained from three independent experiments for ECcon-MVs group and ECogd-MVs group; Vehicle vs ECogd-MVs: p < 0.0001; ECcon-MVs vs ECogd-MVs: p = 0.0012. g Morphologic changes of same AC treated with ECogd-MVs at different time points. h Trans-differentiated cells in (g) were fixed at day 7 post ECogd-MVs treatment; Tuj1, Map2 and GFAP were detected by immunostaining. i Representative images of time-lapse live imaging analyses of individual AC after ECogd-MVs treatment. j Converted cells transduced with hSyn.GCaMP6S AAV and treated with 1 uM tetrodotoxin (TTX). k GCaMP6s traces for regions of interest 1–5; Arrows indicate the time points in (j). l Number of GCaMP+ cells in AC treated with vehicle or ECogd-MVs; n = 3 biologically independent samples for vehicle group, n = 4 biologically independent samples for ECogd-MVs group; p = 0.0217. Experiment was repeated independently 3 times with similar results (c, d, g–j). Data are shown as mean ± SEM. Unpaired two-tailed t-test (a, b, l); one-way ANOVA with post-hoc Tukey adjustment (f). Scale bar, 25 μm (c, d); 50 μm (g–j). Source data are provided as a Source Data file.

Fig. 3. Brain endothelial cells-derived microvesicles containing pro-neural transcription factor Ascl1 initiate astrocytes trans-differentiation.

a Representative immunostaining of Ascl1 and GFAP at day 0, day 1 and day 2 post AC co-cultured with OGD-stimulated brain endothelial cells. b Protein levels of Ascl1 in AC or AC at day 2 post co-cultured with OGD-stimulated brain endothelial cells (ECogd); n = 4 biologically independent samples; p = 0.0173. c mRNA levels of Ascl1 in AC at day 0, day 1 and day 2 post co-cultured with OGD-stimulated brain endothelial cells; n = 3 biologically independent samples; p < 0.0001. d mRNA levels of Ascl1 in normal or OGD-stimulated brain endothelial cells; n = 4 biologically independent samples; p = 0.0471. e Protein levels of Ascl1 in MVs derived from normal or OGD-stimulated brain endothelial cells; n = 4 biologically independent samples; p = 0.0065. f Confocal microscopy revealed co-localization of GFP and Alix in brain endothelial cells transduced with GFP-Ascl1 lentivirus. g Experimental schematic for testing effects of MVs derived from brain endothelial cells overexpressed Ascl1 (ECas1-MVs) on AC. h Luciferase activity showed increased activities of Atoh8 and Hes6 in AC treated with ECas1-MVs; n = 4 biologically independent samples for Atoh8 group, n = 3 biologically independent samples for Hes6 group; p = 0.0265 for Atoh8; p = 0.0009 for Hes6. i Morphologic changes of same AC at day 0, day 1, day 3 and day 5 post ECas1-MVs treatment. j Representative immunostaining of DCX and Nestin in AC at day 3 post ECas1-MVs treatment. k Representative immunostaining of Map2 and Tuj1 in AC at day 7 post ECas1-MVs treatment; BF: bright filed. Experiment was repeated independently 3 times (a, f, i–k) or 4 times (b, e) with similar results. Data are shown as mean ± SEM. Unpaired two-tailed t-test (b, d, e); one-way ANOVA with post-hoc Tukey adjustment (c); two-way ANOVA with post-hoc Bonferroni adjustment (h). Scale bar, 50 μm. Source data are provided as a Source Data file.

Fig. 4. Pro-neural transcription factor Ascl1 is expressed in brain microvessels after focal cerebral ischemia in mice.

Representative immunostaining of Ascl1, CD31, and NeuN at shame (a), day 1 (b), and day 3 (c) post permanent distal middle cerebral artery occlusion (MCAO); Dashed white line indicates lesion border. d Vessels were counted in 0.5 mm² peri-infarct ROIs per section (totally 3 sections per mouse, 4 mice per group; Criteria is any loci with more than one Ascl1 positive cell); p = 0.0027. e Isolation of brain endothelial cell, neuron, pericyte, and microglia/Mφ from ipsilateral cortex by fluorescence-activated cell sorting (FACS) at 2 h post permanent distal MCAO. f mRNA levels of Ascl1 in different cell isolated from (e). The experiment was repeated independently 4 times with similar results (a–c). Data are shown as mean ± SEM. One-way ANOVA with post-hoc Tukey adjustment. Scale bar, 200 μm. Source data are provided as a Source Data file.

Fig. 5. Microvesicles derived from OGD-stimulated brain endothelial cells convert astrocytes into neural progenitor cells in mouse focal cerebral ischemia.

Confocal microscopy revealed Cy3 labeled primary brain endothelial cells-derived microvesicles (EC-MVs) could be taken into astrocytes at 3 h (a, b) and 24 h (a, c) post lateral ventricle injection; Dashed white line indicates lateral ventricle border. d Experimental design. e Representative immunostaining of GFP, GFAP, and NeuN at day 14 post lateral ventricle injection of AAV.GFA104.PI.eGFP. f Representative immunostaining of GFP and DCX at day 10 post 45 mins transient MCAO. g Number of DCX+ cells in 1 mm² peri-infarct ROIs per section (totally 3 sections per mouse, 4 mice per group); p = 0.0026. h Number of both DCX+ and GFP+ cells in 1 mm² peri-infarct ROIs per section (totally 3 sections per mouse, 4 mice per group); p = 0.0043. Experiment was repeated independently 3 times (a) or 4 times (e, f) with similar results. Data are shown as mean ± SEM. Unpaired two-tailed t-test. Scale bar, 25 μm (a); 200 μm (e, f). Source data are provided as a Source Data file.

Fig. 6. Overexpression of Ascl1 in brain endothelial cells increases neurogenesis and improves neurological recovery after focal cerebral ischemia.

a AAV-FLEx-ASCL1-GFP or AAV-FLEx-GFP was injected into lateral ventricle of the Tie2-cre mice. Created with BioRender.com. b Representative immunostaining of GFP, CD31 and NeuN in cortex at day 8 post AAV-FLEx-ASCL1-GFP injection. c Representative immunostaining of GFP, CD31 and GFAP in cortex at day 8 post AAV-FLEx-ASCL1-GFP injection. d Experimental design. e Representative immunostaining of Ascl1 and TUJ1 at day 14 post permanent distal MCAO. f Quantification of TUJ+ cells expressing Ascl1 in the peri-infarct area (totally 3 sections per mouse, 4 mice per group); p = 0.0392. g Quantification of TUJ1+ cells in the peri-infarct area (totally 3 sections per mouse, 5 mice per group); p = 0.0496. h Representative immunostaining of Ascl1 and NeuN at day 14 post distal permanent MCAO; Dashed white line indicates lesion border. i Representative immunostaining of Ascl1 and NeuN at day 14 post distal permanent MCAO after AAV-FLEx-ASCL1-GFP injection. j–k Quantification of NeuN+ cells (j) and Ascl1+ cells (k) in the peri-infarct area (totally 3 sections per mouse, 4 mice per group); p = 0.0245 in (j); p = 0.0323 in (k). l Foot fault test (n = 8 mice for AAV-FLEx-GFP group, n = 9 mice for AAV-FLEx-ASCL1-GFP group); p = 0.0439. The experiment was repeated independently 5 times with similar results (b, c). Data are shown as mean ± SEM. Unpaired two-tailed t-test (f, g, j, k); two-way ANOVA with post-hoc Bonferroni adjustment (l). Scale bar, 25 μm (e, i); 100 μm (b, c, h). Source data are provided as a Source Data file.