Supplementary data related to Fig. 5. (A) Quantification of the mean area of ASO-positive vesicles in either Epha2^+/+ (blue bars) or Epha2^−/− (magenta bars) KPC cells, after treatment with either 0.5, 1, or 5 μM of cET-ASO^Kras for 16 h. One-way ANOVA, Tukey multiple comparison test, n = 7. (B) Quantification of the mean area size of galectin-9–positive vesicles in either Epha2^+/+ (blue bars) or Epha2^−/− (magenta bars) KPC cells, after treatment with either 0.5, 1, or 5 μM of cET-ASO^Kras for 16 h. One-way ANOVA, Tukey multiple comparison test, n = 11. (C) Quantification of the number of ASO vesicles per area of cytoplasm (left panel) and galectin-9 fluorescence intensity (right panel), in cells treated with either vehicle or bafilomycinA1 (10 nM) for 4 h in combination with cET-ASO^Kras (5 μM). n = 3 independent experiments, ANOVA, Tukey post hoc test. (D) 3D-reconstruction of galectin-9 (red) overlap with cET-ASO^Kras (magenta) and G3BP1-GFP (green) in Epha2^+/+ or Epha2^−/− KPC cells. (E and F) Average sum of the area of G3BP1-positive structures per cell (E) and mean area of individual G3BP1 vesicles (F) in either Epha2^+/+ or Epha2^−/− KPC cells after incubation with cET-ASO^Kras for 0, 2, or 16 h. Dots correspond to different fields of view obtained from three individual experiments. One-way ANOVA—Tukey test. (G) Scatter plot of the mean intensity values for G3BP1, ASO, and galectin-9 fluorescence for each cell in either Epha2^+/+ (left panel) or Epha2^−/− (right panel) KPC cells after incubation with cET-ASO^Kras for 16 h. (H) Micrographs of immunofluorescence confocal imaging showing the overlap between G3BP1-GFP, EPHA2, and galectin-9 in intracellular structures in KPC cells treated with cET- ASO^Kras for 16 h. (I) Micrographs of confocal immunofluorescence imaging showing the overlap between G3BP1-GFP, internalized EPHA2^pS897, and intracellular CD44 in KPC cells treated with cET-ASO^Kras for 16 h. Image zooms correspond to the area inside the corresponding dotted line boxes. (J) Micrographs of confocal immunofluorescence imaging showing the overlap between G3BP1-GFP, galectin-9, and eIF3b fluorescence in intracellular structures in KPC cells treated with cET-ASO^Kras for 16 h. (K) Western blotting of either nontargeting (EPHA2^CRISPR-NT) or EPHA2^CRISPR-KO H1299 cells showing the effect on eIF2α phosphorylation after incubation with ASO cET-ASO^Kras for 4 or 6 h. Vinculin was used as a loading control. (L) 3D reconstruction of immunofluorescence confocal imaging showing the overlap between G3BP1-GFP (green), internalized cET-ASO^Kras (red), and intracellular CD44 (pink) in either Epha2^+/+, Epha2^−/−, or ISRIB (1 μM)-treated Epha2^+/+ KPC cells. (M) Western blotting showing the protein expression of G3BP1 in nontargeting (n.t.) and KPC cells transduced with four different sgRNA-targeting G3bp1 in KPC cells (G3bp1^CRISPR-KO, upper left panel); western blotting showing the protein expression of G3BP2 in nontargeting (n.t.), G3bp1^CRISPR-KO#1 and G3bp1^CRISPR-KO#1 KPC cells transduced with three different sgRNAs targeting G3bp2 (G3bp1/2^CRISPR-KO, upper right panel); western blotting showing the protein expression of G3BP1 in nontargeting (n.t.), G3bp1^CRISPR-KO#1, and G3bp1^CRISPR-KO#1 KPC cells transduced with three different sgRNAs targeting G3bp2 (G3bp1/2^CRISPR-KO, lower left panel). GAPDH was used as a loading control. (N) Diagram depicting the endocytic pathway driven by CD44 and EphA2 transporting ASOs from the extracellular space into the perinuclear region and enabling their release from nuclear-captured leaky endosomes. The left panel depicts the situation when both EPHA2 and CD44 levels are low, and there is little internalization of ASO. The center panel summarizes events when CD44 and EPHA2 levels are both high. ASO directly engages with CD44, which then drives activation of p90RSK to phosphorylate EPHA2 on Ser⁸⁹⁷. This promotes internalization of ASO into EPHA2/CD44 positive vesicles, which are then trafficked to the juxta-nuclear region in a Rab17-dependent manner. ASO-containing endosomes are then captured on the nuclear surface via an interaction between the cytotail of EPHA2 and the nuclear pore complex. ASO-driven increase in ROS production and the resulting LP leads to endosomal membrane damage. Subsequently, these vesicles become leaky, allowing escape of the ASO, despite the activation of repair mechanisms under the ISR, such as SG formation. When SG assembly is inhibited using ISRIB, this repair process is compromised, and nuclear-captured endosomes become even more leaky, as denoted in the right panel. Source data are available for this figure: SourceData FS3.