Scalable Production of Glioblastoma Extracellular Vesicles Using Long-term Bioreactor Cultures

Introduction:: Glioblastoma (GBM) is a highly aggressive and lethal form of brain cancer, resulting in both low survival durations and low quality of life for patients. Extracellular vesicles (EVs) are micro- and nanoscale lipid-enclosed particles that are released by all cells and circulate in all bodily fluids, and recent evidence indicates that the various types of EVs released by GBM cells play a crucial role in the progression of the disease and could even serve as liquid biopsy biomarkers or provide insight into the underlying mechanisms driving its lethality. However, EVs can be difficult to produce in bulk using conventional in vitro cultures, severely limiting technology development that requires a large supply of relatively consistent EVs. Commercially available bioreactor systems that were originally designed for antibody production may provide a simple and effective means of maintaining high-density cell cultures for EV production, but these systems have not been explored for scalable GBM EV production to date. In this study, we hypothesize that CELLine AD 1000 bioreactors can be used to support long-term high-density cultures from several GBM cell lines for EV production, and that they outperform conventional cultures based on EV yield, single-use plastic consumption, and usability.

Materials and Methods:: Several cell lines representing various subtypes of GBM (U87-EGFRvIII - classical, U87 - mesenchymal, A172 - neural, U373 – proneural, and BT169 - gliosarcoma) were cultured, and 25 million cells from each cell line were inoculated into the lower 15 ml chamber of each CELLine AD 1000 bioreactor in complete DMEM. Over the first four weeks, the 10% fetal bovine serum (FBS) supplementation was gradually replaced with a chemically defined serum replacement (CDM-HD, Fibercell) to avoid contamination from exogenous FBS EVs and a feeding/harvest schedule of three times per week for the lower 15ml cell chamber and once per week for the upper 1 L nutrient chamber was established. Large and small EVs from the conditioned media in the 15 ml bioreactor chamber, as well as conditioned media from conventional cultures of U87 and U373 cell lines, were isolated (Fig 1A) using combinations of differential centrifugation, tangential flow and ultrafiltration, and size exclusion chromatography (SEC), and EVs were characterized according to MISEV guidelines by western blot, transmission electron microscopy (TEM), bicinchoninic acid assay, and nanoparticle tracking analysis (NTA).

Results, Conclusions, and Discussions:: Initial results indicate that all GBM cell lines tested successfully established high-density cultures within CELLine AD 1000 bioreactors and enabled the scalable and long-term production of GBM EVs for up to four months, producing over two hundred EV preparations total. Small EVs were successfully isolated from soluble protein contaminants by SEC as demonstrated by performing NTA and BCA measurements on individual SEC fractions (Fig 1B). The yield of small EVs from early U87 and U383 bioreactor cultures was equivalent to at least 10 x T175s using conventional culture methods, and importantly did not require tangential flow filtration to concentrate the large volume of conditioned media (15 ml bioreactor vs 300 ml conventional). Additionally, U87, U373, and BT169 bioreactors outperformed A172 and U87-EGFRvIII bioreactors during early cultures. Small EVs from all five bioreactors were validated according to MISEV guidelines, showing expected size distributions by NTA and cup-like morphology by negative staining TEM, as well as the presence of exosome-associated tetraspanin CD81 and lack of endoplasmic reticulum-associated GRP94 by western blotting (Fig 1C).



Overall, this scalable EV production and isolation workflow adds to the existing body of evidence demonstrating that commercially available bioreactors can produce significantly more EVs than conventional cultures with less single-use plastic waste, without the need for passaging, and without contamination by exogenous EVs from serum supplementation. As the EV field continues to grow, this approach will enable the development of many technologies by providing vast amounts of EVs for applications ranging from liquid biopsies and therapeutics to mechanistic understandings of disease progression.


Acknowledgements (Optional): : This work was funded by The Ohio State University’s CCE-CURES Accelerator Award, the LEGACY Postdoctoral Scholars Program, and the National Institutes of Health (NIBIB - K99EB033857).

References (Optional): :