Glioblastoma is a primary malignant brain tumor characterized by highly infiltrative glioma cells. Vasculature and white matter tracts are considered to be the preferred and fastest routes for glioma invasion through brain tissue. In this study, we systematically quantified the routes and motility of the U251 human glioblastoma cell line in mouse brain slices by multimodal imaging. Specifically, we used polarization-sensitive optical coherence tomography to delineate nerve fiber tracts while confocal fluorescence microscopy was used to image cell migration and brain vasculature. Somewhat surprisingly, we found that in mouse brain slices, U251 glioma cells do not follow white matter tracts but rather preferentially migrate along vasculature in both gray and white matter. In addition, U251 cell motility is ∼2-fold higher in gray matter than in white matter (91 vs. 43 μm2/h), with a substantial fraction (44%) of cells in both regions invading without close association with vasculature. Interestingly, within both regions, the rates of migration for the perivascular and televascular routes of invasion were indistinguishable. Furthermore, by imaging of local vasculature deformation dynamics during cell migration, we found that U251 cells are capable of exerting traction forces that locally pull on their environment, suggesting the applicability of a “motor-clutch”-based model for migration in vivo. Overall, by quantitatively analyzing the migration dynamics along the diverse pathways followed by invading U251 glioma cells as observed by our multimodal imaging approach, our studies suggest that effective antiinvasive strategies will need to simultaneously limit parallel routes of both perivascular and televascular invasion through both gray and white matter.
Bibliographical noteFunding Information:
The study was supported by research grants from National Institutes of Health ( R01-CA-172986 and U54-CA-210190 to D.J.O.) and by the Graduate School Doctoral Dissertation Fellowship at the University of Minnesota (to C.J.L.). High-performance computing resources were provided by Minnesota Supercomputing Institute.
The authors thank Xiyao Jin for assistance with data collection and Aleksi Isomursu for help with the cell migration simulator. The study was supported by research grants from National Institutes of Health (R01-CA-172986 and U54-CA-210190 to D.J.O.) and by the Graduate School Doctoral Dissertation Fellowship at the University of Minnesota (to C.J.L.). High-performance computing resources were provided by Minnesota Supercomputing Institute.
© 2019 Biophysical Society