A2 prevented BDEV-induced vascular leakage. (A) Water content of the brains from TBI mice receiving A2 or TBST, showing a dose-dependent reduction of brain water content in A2-treated TBI mice (red dots), compared with that of untreated mice (white dot, n = 12, 1-way ANOVA, P < .01 vs untreated mice as control). (B-I) Laser speckle images from TBI mice treated with either A2 or TBST (n = 8). (J) Blood flow of the contusional and pericontusional cortex in the region of interest (ROI) from TBI mice receiving either A2 or TBST and sham mice (n = 9, 1-way ANOVA, *P < .01 among 3 groups at 3 and 24 hours, **P < .01 vs baseline in TBI-A2 mice, ***P < .001 vs baseline in TBI-TBST mice). (K-M) HUVECs were cultured to confluence and then treated with plasma from sham and TBI mice (K) receiving TBST (L) or A2 (M) for 3 hours at 37°C. The plasma samples were obtained 3 hours post-TBI. The cells were stained with a CD31 antibody (red) to mark the tight junction (scale bars, 100 μm; arrows: disrupted junctions, representative of 3 separate experiments). (N) PKH26-labeled EVs purified from equal volumes of plasma from sham mice and TBI mice treated with A2 or TBST were incubated with HUVECs for 3 hours at 37°C in the presence of 3 × 10⁵/μL of human platelets. The labeled EVs were quantified in the bottom chamber using flow cytometry (n = 12, 1-way ANOVA). (O) The same experiments as described in panel N were used to measure the transendothelial migration of PKH26-labeled BDEVs (n = 12, 1-way ANOVA).