Fig 2: Mitochondrial injury drives redox-mediated phase separation of ApoE2 and ApoE3.(A) Percentage of ApoE condensates with volumes greater than 0.6 μm3 in RPE expressing ApoE2, ApoE3, or ApoE4 and treated with or without A2E or NHS. (B) IUPRED2A disorder plots for human ApoE2, ApoE3, and ApoE4. Redox state–dependent order-disorder transitions are depicted by the pink shaded area. Note increased disorder corresponding to cysteines at 112 and 158 in ApoE2 and at 112 in ApoE3. (C) Frequency distribution of condensate volumes in ApoE2-expressing RPE with A2E exposed to NHS and treated with simvastatin (5 μM, 16 h), T0901317 (1 μM, 16 h), or desipramine (10 μM, 3 h). Mean ± SEM, *P < 0.05, **P < 0.005, ***P < 0.0001. n = 15–29 cells per condition; 1-way ANOVA with Bonferroni’s posttest. See also Supplemental Figure 4.

Fig 3: ApoE isoform-specific differences in regulating RPE cholesterol transport and microtubule dynamics.(A) Stills from live imaging of mCherry-tagged human ApoE2, ApoE3, or ApoE4 (top panel, red) in primary RPE cultures. Tracks of individual ApoE vesicles (lower panel). Color bar shows displacement of individual tracks from short (cooler colors) to long (warmer colors); range, 0.0 µm to 4.212 µm. (B) Percentage of total ApoE tracks with displacement > 0.8 µm in RPE expressing ApoE2, ApoE3, or ApoE4. Mean ± SEM, n >18 cells per condition from 3 independent experiments. **P < 0.005. (C) Endogenous cholesterol content in mock-transfected or ApoE2-, ApoE3-, or ApoE4-expressing RPE treated with A2E. Mean ± SEM from 3 independent experiments, 3 replicates/experiment. (D) Representative images and quantitation of acetylated tubulin (green) immunostaining in mock-transfected or ApoE2-, ApoE3-, or ApoE4-expressing RPE treated with or without A2E. Cell boundaries are demarcated by zonula occludens 1 (ZO-1) (white) and nuclei are labeled with DAPI (blue). Cells expressing mCherry-ApoE (red) are outlined in yellow. (E) Quantification of acetylated tubulin intensity in mCherry-expressing cells. Mean ± SEM, n = 10 cells/condition. (B, C, and E) One-way ANOVA with Bonferroni’s posttest. *P < 0.01, **P < 0.001, ****P < 0.0001. See also Supplemental Figure 1.

Fig 4: Mitochondrial injury and redox-sensitive protein phase separation in Abca4–/– mouse RPE.(A) Volume reconstructions of TOM20-stained mitochondria in WT and Abca4–/– mouse RPE flatmounts. Warmer colors in the color bar indicate larger volumes. (B) Quantification of mean TOM20 volumes from images in A. Mean ± SEM, *P < 0.05, n = 3 mice per genotype. (C) ApoE immunostaining (green) in WT and Abca4–/– mouse RPE flatmounts. (D) IUPRED2 redox-driven disorder predictions for mouse p62. (E) p62 condensates (green) in WT and Abca4–/– RPE flatmounts. (F) Areas of p62 aggregates. Mean ± SEM; *P < 0.05; n = 3 mice per genotype. (G) p62 (green) in RPE in Abca4–/– mice treated with vehicle or desipramine. (C, E, and G) Nuclei, DAPI (blue); actin, phalloidin (gray).

Fig 5: ApoE2 exacerbates complement-induced mitochondrial fragmentation in the RPE.(A) 3D reconstruction of mitochondrial volumes from live imaging of MitoTracker in mock-transfected RPE or RPE expressing mCherry-tagged ApoE2, ApoE3, or ApoE4 exposed to 10% NHS to induce complement attack. Cooler colors in the color bar indicate increasing mitochondrial fragmentation. (B) Number of fragmented mitochondria after NHS exposure. (C) Mitochondrial volumes as in A in RPE treated with A2E prior to NHS exposure. (D) Quantification of mitochondrial fragments as in B. (E) Mitochondrial volumes as in A in ApoE2-expressing RPE with A2E exposed to NHS and treated with simvastatin (5 μM, 16 h), T0901317 (1 μM, 16 h), or desipramine (10 μM, 3 h) prior to imaging. (F) Quantification of mitochondrial fragments as in B. Mean ± SEM, *P < 0.05, **P < 0.005, ***P < 0.0005, ****P < 0.0001. n = 30 cells (A), 26 cells (B), and 30 cells (C) per condition; 1-way ANOVA with Bonferroni’s posttest. See also Supplemental Figure 3.