Red Blood Cell Mechanical Stimulation and Storage Triggers Vasoactive Extracellular Vesicle Release

Introduction: Red blood cells (RBCs) are a promising source of extracellular vesicles (EVs) for large-scale biomanufacturing because they are easily accessible, require lower maintenance than nucleated cells, and can produce high EV yield. RBC-EVs are thought to be safe because RBCs have simpler structure and function compared to other cell types. However, RBCs release biochemical signals that affect the function of the endothelium, the innermost lining of all blood vessels that regulate vascular hemostasis. However, it is unclear whether these signals are packaged into RBC-EVs or how RBC-EV contents affect vascular function. Therefore, this study aimed to determine how RBC-EV biomanufacturing methods affect RBC-EV contents and subsequent endothelial function.

RBC-EVs can be harvested from stored RBCs; however, storage RBC-EVs are considered atherogenic because they contain hemoglobin, which can promote endothelial dysfunction. We recently showed that activating RBC piezo1, a mechanosensitive channel that increases RBC Ca2+ influx, triggers RBC-EV release within 30 minutes. Piezo1 stimulation rapidly generates RBC-EVs from freshly isolated RBCs. Additionally, piezo1 RBC-EVs may not promote vascular dysfunction, considering they model physiologically relevant RBC-EVs released as RBCs deform through microcirculatory transit. Therefore, we hypothesized that piezo1 RBC-EVs from freshly isolated RBCs would not promote endothelial dysfunction compared to RBC-EVs released from stored RBCs.

Materials and

Methods: RBCs isolated from freshly drawn whole blood (six people) using protocols approved by the University of Maryland Institutional Review Board (IRB Net ID: 1385293-7) and conformed to the Declaration of Helsinki. RBCs (6% hematocrit) were resuspended in citrate-phosphate-dextrose-adenine solution and treated with 10 µM piezo1-agonist yoda1 for 30 minutes or stored at 4oC for one month. RBC-EVs were isolated using differential ultracentrifugation and then characterized using nanoparticle tracking analysis (NTA) to quantify RBC-EV size and concentration; Western blot to detect EV markers ALIX, TSG101, CD9, and RBC marker stomatin; and transmission electron microscopy (TEM) to visualize RBC-EV morphology. Untargeted liquid chromatography-mass spectrometry (LC-MS) was used to compare piezo1 and storage RBC-EV contents. Finally, we used a scratch assay to determine how piezo1 and storage RBC-EVs affect endothelial migration. Human umbilical vein endothelial cells (HUVECs) were cultured in endothelial growth media-2 (EGM-2) supplemented with 1% penicillin-streptomycin and 1% glutamine. HUVEC were then scratched with a one-milliliter pipette tip and treated with piezo1 or storage RBC-EVs (e10 particles/mL) for 24 hours. Untreated HUVEC were used as controls. Finally, untreated and RBC-EV-treated HUVEC were imaged at 0, 4, 8, 12, 16, and 24 hours to quantify how RBC-EVs affect endothelial migration.

Results, Conclusions, and Discussions: RBCs treated with piezo1-agonist yoda1 released EVs that were 178±13 nm in diameter with a yield of 7.3xe8 particles/mL, while RBC storage generated EVs that were 164±15 nm in diameter with a yield of 4.8xe8 particles/mL (Fig 1a). We verified isolation of RBC-EVs using TEM, which revealed that RBCs released biconcave shaped vesicles on the nanometer scale (Fig 1a inset). Western blot showed that both piezo1 and storage RBC-EVs contained EV markers ALIX and CD9 and RBC marker stomatin, but only piezo1 RBC-EVs contained EV marker TSG101 (Fig 1b). LC-MS revealed that piezo1 and storage RBC-EVs contained 2560 different metabolites (Fig 1c). Finally, we found that storage RBC-EV decreased endothelial cell migration by 12 hours compared to endothelial cells treated with piezo1 RBC-EVs. At the 24-hour timepoint, storage RBC-EVs had significantly decreased endothelial migration by 44.6% compared to endothelial cells treated with piezo1 RBC-EVs (p < 0.05; Fig 1d).

Our study demonstrates that RBC-EVs manufactured under different conditions have different contents and impact on endothelial function. Specifically, RBC-EVs generated from piezo1 stimulation support endothelial migration compared to those obtained from prolonged storage, which diminish cell migration. Thus, piezo1 RBC-EVs are more likely to maintain endothelial function. This study is impactful because our data also indicates that RBC-EVs carry contents other than hemoglobin that can potentially impair endothelial migration. Future work will determine how storage RBC-EVs mechanistically impair endothelial migration and investigate how piezo1 and storage RBC-EVs affect other hallmarks of endothelial dysfunction. Understanding how RBC-EV manufacturing methods affect endothelial function will enable safe biomanufacturing protocols to create RBC-EV-based drug delivery platforms that do not promote vascular dysfunction.

Acknowledgements (Optional): Funding was provided to G.S.S by the AHA Postdoctoral Fellowship (916512) and A.M.C by NIH (R01HL140239; R01HL165193).