BACKGROUND: The physiopathology of life-threatening cerebrovascular complications in preeclampsia is unknown. We investigated whether disruption of the blood-brain barrier, generated using circulating small extracellular vesicles (sEVs) from women with preeclampsia or placentae cultured under hypoxic conditions, impairs the expression of tight junction proteins, such as CLDN5 (claudin-5), mediated by VEGF (vascular endothelial growth factor), and activation of KDR (VEGFR2 [VEGF receptor 2]). METHODS: We perform a preclinical mechanistic study using sEVs isolated from plasma of pregnant women with normal pregnancy (sEVs-NP; n=9), sEVs isolated from plasma of women with preeclampsia (sEVs-PE; n=9), or sEVs isolated from placentas cultured in normoxia (sEVs-Nor; n=10) or sEVs isolated from placentas cultured in hypoxia (sEVs-Hyp; n=10). The integrity of the blood-brain barrier was evaluated using in vitro (human [hCMEC/D3] and mouse [BEND/3 (brain endothelial cell 3)] brain endothelial cell lines) and in vivo (nonpregnant C57BL/6J mice [4-5 months old; n=13] injected with sEVs-Hyp) models. RESULTS: sEVs-PE and sEVs-Hyp reduced total and membrane-associated protein CLDN5 levels (P<0.05). These results were negated with sEVs-PE sonication. sEVs-Hyp injected into nonpregnant mice generated neurological deficits and blood-brain barrier disruption, specifically in the posterior area of the brain, associated with brain endothelial cell uptake of sEVs, sEVs-Hyp high extravasation, and reduction in CLDN5 levels in the brain cortex. Furthermore, sEVs-PE and sEVs-sHyp had higher VEGF levels than sEVs-NP and sEVs-Nor. Human brain endothelial cells exposed to sEVs-PE exhibited a reduction in the activation of KDR. Reduction in CLDN5 observed in cells treated with sEVs-Hyp was further enhanced in cells treated with KDR selective inhibitor. CONCLUSIONS: sEVs-PE disrupts the blood-brain barrier, an effect replicated with sEVs-Hyp, and involves reduced CLDN5 and elevated VEGF contained within these vesicles. However, our results do not support the participation of KDR activation in the downregulation of CLDN5 observed with sEVs-Hyp. These findings will improve our understanding of the pathophysiology of cerebrovascular alterations in women with preeclampsia.