Article
Six extracellular vesicle related genes can explain the pro-tumorigenic behavior of heterogeneous high grade gliomas
Search Medline for
Authors
-
Franz Ricklefs
- Universitätsklinikum Hamburg-Eppendorf, Klinik für Neurochirurgie, Brigham and Women's Hospital, Harvard Medical School, Depratment of Neurosurgery, Hamburg, Deutschland
-
Marco Mineo
- Brigham and Women's Hospital, Harvard Medical School, Department of neurosurgery, Boston, United States
-
Arun Rooj
- Brigham and Women's Hospital, Harvard Medical School, Department of neurosurgery, Boston, United States
-
Jakub Godlewski
- Brigham and Women's Hospital, Harvard Medical School, Department of neurosurgery, Boston, United States
-
Ennio Antonio Chiocca
- Brigham and Women's Hospital, Harvard Medical School, Department of neurosurgery, Boston, United States
-
Manfred Westphal
- Universitätsklinikum Hamburg-Eppendorf, Klinik für Neurochirurgie, Hamburg, Deutschland
-
Agnieszka Bronisz
- Brigham and Women's Hospital, Harvard Medical School, Department of neurosurgery, Boston, United States
| Published: | June 9, 2017 |
|---|
Outline
Text
Objective: High grade gliomas display cellular hierarchical heterogeneity descending from a subpopulation of glioma stem-like cells (GSCs). The lack of experimental models hinders our understanding of intratumoral interactions and therefore cannot reflect the complex subpopulation dynamics within the tumor microenvironment. Extracellular vesicles (EVs) shed by tumor cells have recently been identified to serve as major conduits in cell-cell communication but have not been studied in a subtype specific manner.
Methods: 8 patient-derived GSCs were chosen after clustering to either the mesenchymal (M) or proneural (P) subtype. Bioinformatic analysis was used to examine mass spectra profiles of GSC EVs followed by validation via immunoblotting of cell culture and patient serum EVs. EV release and transfer was visualized by stable PALM-dtTomato M GSC or PALM-GFP P GSC and subsequent confocal microscopy imaging. Changes in gene expression profiles were tested by RT-PCR. Migration of P and M GSCs was monitored in vitro and in vivo.
Results: We identified over 1400 proteins within the EV proteome which shown subtype-dependent signature with different biological and molecular modules. Retrospective data extrapolated from the TCGA database revealed that 3 genes of either M or P subtype-specific EV proteome were associated with worse outcome in the other patient subtype. Accordingly, transfer of P and M EVs and their protein cargo led to increased pro-tumorigenic traits in vitro and in vivo with a subtype-specific uptake pattern and migratory/invasion phenotype. The subset of P EV proteins were identified as promoting glioma invasiveness, and cancer promoting molecules which build the network pathways with its cellular subtype specific effectors.
Conclusion: The secretion and uptake of EV mediates dynamic intra-tumoral interactions between cells from the same and different subtypes, providing a foundation for understanding how tumor heterogeneity affects patients’ outcomes. Furthermore specific protein composition might enable detection of tumor derived EVs that could work as biomarker.
