(J-388) Differences in angiogenic and immunomodulatory capacities of extracellular vesicles derived from bone marrow- and adipose-derived mesenchymal stromal/stem cells and impacts of in vitro expansion

Introduction. Recently, therapies utilizing extracellular vesicles (EVs) derived from mesenchymal stromal/stem cells (MSCs) have begun to show promise in clinical trials. However, EV therapeutic potential varies with MSC tissue source and in vitro expansion through passaging. To find the optimal MSC source for clinically translatable EV-derived therapies, this study aims to compare the angiogenic and immunomodulatory potentials and the protein and miRNA cargo compositions of EVs isolated from the two most common clinical sources of adult MSCs, bone marrow and adipose tissue, across different passage numbers.
Material and Methods. Primary bone marrow-derived MSCs (BMSCs) and adipose-derived MSCs (ASCs) were isolated from adult female Lewis rats and expanded in vitro to the indicated passage numbers (P2, P4, and P8). MSC surface markers were confirmed by flow cytometry. EVs were isolated from the culture medium of P2, P4 and P8 BMSCs and ASCs. EV size, number, surface markers, protein content, and morphology were characterized via nanoparticle tracking analysis, western blotting, BCA kit, and transmission electron microscopy, separately. EV immunomodulatory and angiogenic capacity were assessed via qRT-PCR, ELISA, cell proliferation assays, and tube formation assays employing RAW264.7 cells, human umbilical vein endothelial cells (HUVECs), and primary BMSCs. Proteomics and small RNA sequencing were performed for EV cargo profiling.
Results, Conclusions, and Discussions. Although both MSC sources secreted EVs with similar mean sizes and average amounts of protein at all passage numbers, BMSCs showed a significantly higher EV yield per cell than ASCs at P2 (p < 0.01, n > 6) and P4 (p < 0.05, n > 6), but not at P8 (n > 3). The protein content per EV also declined as the passage number increased in both tissue sources. Transmission electron microscopy and western blotting confirmed the morphology and surface marker expression of all EVs studied. Proteomics analysis indicated that BMSC-EVs showed a higher number of unique protein types than ASC-EVs at the same passage, while BMSC-EVs and ASC-EVs each showed distinct protein profiles at different passage numbers. In vitro tube formation assays employing HUVECs suggested that both tissue source and passage number had significant effects on the angiogenic capacity of EVs (n ≥ 4). Interestingly, proliferation assays showed a higher HUVEC proliferation rate when co-cultured with BMSC-EVs than with ASC-EVs at each passage (n = 3). With or without LPS stimulation, EVs more significantly impacted expression of M2-macrophage genes (IL-10, Arg1, TGFβ) than M1-macrophage genes (IL-6, NOS2, TNFα). The viability of RAW264.7 cells after EV treatment slightly decreased compared to controls, however, this difference did not reach statistical significance at each passage (n = 3). P8 BMSC-EVs increased primary BMSC proliferation, while P2 and P4 BMSC-EVs showed lesser effects. By contrast, ASC-EVs decreased BMSC proliferation at all passage numbers, with the most dramatic effects at P2 (n = 3). GO and KEGG pathway analysis of proteomics data and miRNA seq data identified key proteins and pathways associated with: differences between BMSC-EVs and ASC-EVs; differences due to passage number. By correlating these proteomics analyses with the miRNA seq analysis and differences observed in our in vitro immunomodulatory, angiogenic and proliferation assays, we identified the optimal MSC source and key EV proteins and/or miRNAs for production of EVs for tissue regeneration therapies.