Fig. 5

Distinct signaling programs mediate dasatinib resistance and persistence. a t-SNE plot displays single cells from day 28 cultures as in Fig. 2b, colored by their expression score for a Notch signaling gene signature. Inset shows the same plot with cells from the e86vac jackpot lineage highlighted in red. b Boxplots depict the distribution of Notch signature scores across all single cells (n = 2982) or e86vac lineage cells (n = 30). Horizontal line depicts sample median, box delimits the interquartile range, open circles indicate suspected outliers, and whiskers delimit the non-outlier distribution of data (distribution of data falling within range of median ± 1.5(interquartile range)).c Western blot shows expression of IRS2, activated intracellular Notch (Notch ICD), the Notch-associated transcriptional factor RBPJK and Actin. Input samples correspond to drug-naïve GSC8, non-jackpot clones from Experiment #2 and jackpot clones with IRS2 amplification grown in the presence of dasatinib. The non-jackpot “persister” lineages have high Notch ICD and RBPJK levels, while the jackpot lineage that outgrew Experiment #2 has high IRS2 expression. d Barplots compare growth of parental GSC8, non-jackpot clones from Experiment #2, and jackpot clones from Experiment #2 with IRS2 amplification after at least 30 days of culture in the presence of dasatinib. Dasatinib-containing cultures were treated with the indicated concentrations of γ-secretase/Notch inhibitor or DMSO control (**, p < 0.01; *, p < 0.05 by two-tailed Student’s t test; error bars depict the standard error of at least 4 independently treated replicates). The IRS2-amplified lineage is not dependent on Notch signaling. e t-SNE plot displays single cells from day 28 cultures as in Fig. 2b, colored by their expression score for an AKT gene signature. f Boxplots depict the distribution of AKT signature scores across single cells (n = 2982) or e86vac lineage cells (n = 30). Boxplot parameters are as in panel a. g Western blot shows levels of AKT, its active phospho-serine 463 isoform (P-AKT), and tubulin loading control. Input samples are as in panel c. h Barplot with standard error bars depicts densitometric analysis (ImageJ, NIH) of multiple (n = 3) separate Western blot experiments assessing AKT phosphorylation (Phospho-Serine 463) relative to total AKT in dasatinib-exposed cultures. *, p < 0.05 by two-tailed Student’s t test. i Schematic depicts epigenetic and genetic events proposed to confer drug tolerance, resistance, and relapse in glioblastoma. Gray circles represent drug-naïve glioma cells; red circles represent Notch-dependent persisters that tolerate therapy but proliferate slowly; blue circles represent a subclone that has acquired a genetic amplification of the IRS1 or IRS2 locus that drives AKT signaling despite upstream therapeutic inhibition