Adaptive capacities of cancer cells in Glioblastoma. (Doctoral thesis)

Glioblastoma (GBM) is a highly malignant brain tumor where no curative treatment is available. According to the cancer stem cell (CSC) hypothesis GBMs rely on a small subpopulation of cancer cells with stem-like properties responsible for tumor progression and recurrence. Recent experimental data from GBM and other cancers however suggest that CSCs cannot be defined by a specific marker expression and may in fact not be a stable entity but a population of cells adapting to a changing microenvironment. Here we examined inter- and intra-tumoral heterogeneity of GBM cells and their adaptation capacity based on stem cell-associated marker expression profiles. Tumor cell subpopulations were classified based on their expression of four chosen cell membrane markers (CD133, CD15, A2B5 and CD44) using multicolor flow cytometry. 16 subpopulations were separated and analyzed for their self-renewal capacity and their ability to reform the original heterogeneous cell population in different environmental conditions (normoxia, hypoxia, differentiation and in vivo). Mathematical modeling was applied to calculate state transitions between phenotypes and predict the adaptive response of tumor cells. Similar to GBM biopsies, we observed markers to be heterogeneously expressed in glioma stem-like cells and primary cultures. All analyzed tumor cell subpopulations were able to proliferate indefinitely and carried stem-cell properties including self-renewal potential. Moreover, all subpopulations were able to adapt their marker expression profiles to give rise to the original subpopulations. Interestingly, mathematical modeling revealed a different propensity in reforming the original heterogeneity between subpopulations over time, which was independent of their proliferation index. Each environment presented its specific equilibrium with defined proportions of the subpopulations. This equilibrium was reversible upon environmental change for each tested condition. Single cell RNA-seq analysis revealed no transcriptomic differences between subpopulations. Our results suggest that glioma stem-like cells do not represent a stable entity and that intra-tumoral heterogeneity in GBM at least partially results from a high cellular adaptation capacity. This implies that glioma treatment approaches should take into account the strong propensity of cancer cells for phenotypic state transitions.