Background: Ceramides are pro-apoptotic and pro-inflammatory sphingolipids, which have been shown to drive atherosclerotic plaque development as well as cardiovascular event rates.  Mechanistically, ceramide accumulation triggers endothelial apoptosis and oxidative stress. Furthermore, hyperglycemia in diabetes mellitus activates ceramide production in the endothelium. In recent years, large extracellular vesicles (lEVs) have been identified as intercellular transport vehicles with strong implications for cardiovascular disease development. A potential role for lEVs as ceramide transporters and inductors of diabetes-associated endothelial apoptosis has never been investigated.

Methods and Results: A lipidomic analysis of human coronary artery endothelial cells (HCAECs) and their lEVs showed, that hyperglycemia (+30mM glucose for 72 h) increases ceramide levels in lEVs, with C16 ceramide (d18:1-16:0) being the most abundant ceramide (A). Dihydroceramide levels were also found to be increased but on a much lower level of abundance and sphingomyelin levels were unchanged in lEVs after hyperglycemic injury. Furthermore, we found that neutral sphingomyelinase 2 (nSMase-2) expression was elevated (B) in glucose-treated HCAECs while acid sphingomyelinase and neutral sphingomylinase 1 as well as Ceramide Synthase 5 and 6 expression were unchanged. Ceramide rich-lEVs from HCAECs after hyperglycemic injury reduced viability and triggered apoptosis in the lEV-recipient HCAECs (C). The effect was reproduced in vivo, where streptozotocin (STZ) injection was used to induce hyperglycemia in mice. Treatment with lEVs from hyperglycemic mice lead to higher rates of apoptosis in murine endothelial cells compared to lEVs from normoglycemic mice (D). Furthermore, we generated ceramide-rich lEVs via exogenous application of C16 ceramide and could demonstrate ceramide uptake and transfer as well as apoptosis induction in lEV-recipient HCAECs through activation of Caspase 3/7 (E). The increased ceramide release from HCAECs via lEVs after hyperglycemia was counteracted by GW4869 a pharmacological inhibitor of neutral sphingomyelinase (F). GW4869 also led to a reduction of the pro-apoptotic effect of lEV-treatment on HCAECs. A similar effect was achieved by applying RNAi transfection against neutral sphingomylinase 2 (G). Additionally, we tested if hyperosmolar stress induced by mannitol would increase vesicular ceramide export. Mannitol treatment had no effect on nSMase activity or lEV-recipient cell apoptosis. Furthermore, the nSMas2-dependent apoptosis induction was shown to be restricted to lEVs. In regard to small EVs hyperglycemic injury or GW4869 did not affect apoptosis induction in EV-recipient cells.

Conclusion: lEVs mediate the induction of apoptosis in endothelial cells after hyperglycemic injury through a nSMase-2-dependent intercellular transfer of ceramides. This novel finding emphasizes the important role of ceramides in diabetes associated endothelial injury. Furthermore, we establish a new role for lEVs as intercellular transporters of ceramides in the cardiovascular context. Lastly, these data suggest that targeting nSMase-2 is a promising approach to alleviate endothelial apoptosis in patients with diabetes mellitus type 2.