Permeation of small extracellular vesicles across a human blood-brain barrier transwell model remains below particle detection limits, even under oxygen/glucose deprived conditions

Ischemic stroke disrupts blood-brain barrier (BBB) integrity and alters small extracellular vesicle (sEV) signaling, yet the mechanisms underlying sEV transport across compromised barriers remain poorly understood. This study investigated BBB-sEV interactions under normal and stroke-mimicking oxygen/glucose deprivation (OGD) conditions using an in vitro human BBB co-culture model consisting of brain capillary endothelial (BCECs, hCMEC/D3) and astrocytoma cells (1321N1). Model characterization revealed that co-culture with 1321N1 cells enhanced BBB vulnerability to OGD compared to hCMEC/D3 mono-culture. OGD exposure (5 h and 24 h) progressively decreased transendothelial electrical resistance (TEER) and increased FITC-dextran 4 (FD4) permeability, with more severe impairment in co-cultures (e.g. TEER – after 5 h: 0.85-fold, after 24 h: 0.55-fold for co-culture related to mono-culture). Following 19 h oxygen and glucose restoration (‘recovery’) after 5 h OGD, barrier integrity loss was halted but not reversed. Transcriptomic analysis revealed adaptive cellular responses including upregulated glucose transporter 1 (GLUT1) and vascular endothelial growth factor (VEGF), alongside temporal changes in tight junction protein expression (CLDN5, CDLN6). sEV secretion kinetics in apical and basolateral compartments demonstrated that both cell types released particles in response to OGD in a time-dependent manner, with co-cultures showing enhanced secretion compared to mono-cultures. sEV uptake and permeation studies using eight cancer cell line-derived sEVs revealed cell-origin dependent internalization patterns by BCECs, with the highest uptake for HEK293T and SH-SY5Y sEVs. These internalized sEVs were predominantly targeted to lysosomes. Despite severe barrier disruption due to OGD transcellular permeation of single sEV particles was not detectable.