Rapid and reproducible generation of glioblastoma spheroids for high-throughput drug screening

Identifying new substances that could potentially be used for tumor therapy and the precise analysis of their spectrum of action requires models that are as similar as possible to the tumor present in the patient. Traditionally, two-dimensional (2D) cell cultures are used. However, these only resemble solid tumors to a limited extent. More realistic <i>in vivo</i> models, such as tissue cultures, which are invaluable for the final analysis of the effect of new substances, are unsuitable for high-throughput screening (HTS), such as substance library screening. Therefore, we addressed which parameters need to be optimized to produce 3D cultures suitable for HTS using established tumor cell lines and ultra-low attachment plates, and we tested which experimental parameters need to be considered. In our studies, we have focused on cell lines from gliomas. Gliomas are incurable tumors of the central nervous system and are the subject of intensive research. Our studies used ten glioma cell lines from which we generated spheroids using ultra-low attachment plates. We then determined the spheroid size as a function of the initial cell number and the culture time. We analyzed cell viability using propidium iodide staining, evaluated the effects of temozolomide and radiation on spheroids, and compared the effect to that on 2D cultures. We found that spheroid size correlated linearly with the initial cell number. Fewer cells (250-500) generally resulted in better growth than a higher number. However, not all cell lines produced growing spheroids at all. The spheroids had an outer layer of living cells and an inner core of dead cells. The size of the inner core of dead cells was different in the various cell lines and developed differently during the incubation period. Radiation affected spheroids more than 2D cultures, especially at higher cell densities. Our results provide insight into using glioma cell lines to form spheroids as model systems. We have identified initial cell numbers as a critical parameter for their effective use in research, offering a hopeful outlook for tumor therapy research and drug development.