In glioblastoma, the therapeutically intractable and resistant phenotypes can be derived from glioma stem cells, which often have different underlying mechanisms from non-stem glioma cells. Aberrant signaling across the EGFR-PTEN-AKT-mTOR pathways have been shown as common drivers of glioblastoma. Revealing the inter and intra-cellular heterogeneity within glioma stem cell populations in relations to signaling patterns through these pathways may be key to precision diagnostic and therapeutic targeting of these cells. Single cell parallel proteomic heterogeneity profiling of the EGFR-PTEN-AKT-mTOR pathways was conducted in a panel of fifteen glioma stem cell models derived from patient glioblastoma biopsies. The analysis included 59,464 data points from 14,866 cells and identified forty-nine molecularly distinct signaling phenotypes. High content bioinformatics resolved two unique patient clusters diverging on EGFR expression and AKT/TORC1 activation. Phenotypic validation indicated drug responsive phenotypes to EGFR blocking in the high EGFR expressing cluster with lower tumor initiating potential in comparison to the AKT/TORC1 activated cluster. High EGFR expression trended with improved patient prognosis while AKT/TORC1 activated samples trended with poorer patient outcomes. Genetic heterogeneity was observed in both clusters with proneural, classical and mesenchymal subtypes observed. Quantitative single cell heterogeneity profiling reveals divergent EGFR-PTEN-AKT-mTOR pathways of patient derived glioma stem cells, which would inform future research and personalized therapeutic strategies.
Read full abstract