Fig 2: ARF6 GTP-locked mutant decreases retrograde lysosome pausing.(A-B) Example kymographs and quantification of LAMP1-HT motile fractions under the expression of CFP-ARF6WT, ARF6Q67L, or ARF6T27N. n = 10 neurons; two-way ANOVA with Tukey’s multiple comparisons; anterograde (WT v. QL, P = 0.0428; WT v. TN, P = 0.0314; QL v. TN, P = 0.9915); stationary/bidirectional (WT v. QL, P = 0.0011; WT v. TN, P = 0.2466; QL v. TN, P = 0.0960); retrograde (WT v. QL, P = 0.4213; WT v. TN, P = 0.6013; QL v. TN, P = 0.0729). Symbols indicate comparison to ARF6WT. (C-F) Number of seconds paused per min in each of the 3 conditions for all LAMP1 puncta (C; n = 223–274 puncta; WT v. QL, P < 0.0001; WT v. TN, P = 0.0059; QL v. TN, P = 0.0001), LAMP1 with (D; n = 49–67 puncta; WT v. QL, P = 0.0029; WT v. TN, P > 0.9999; QL v. TN, P = 0.0265) and without LC3 (D; n = 174–206 puncta; WT v. QL, P < 0.0001; WT v. TN, P = 0.0198; QL v. TN, P = 0.0431), and all the retrograde- (E; n = 49–65 puncta; WT v. QL, P = 0.0004; WT v. TN, P = 0.6390; QL v. TN, P = 0.0574) or anterograde- (E; n = 27–73 puncta; WT v. QL, P = 0.0507; WT v. TN, P > 0.9999; QL v. TN, P = 0.0846) moving LAMP1 puncta. Kruskal-Wallis test with Dunn’s multiple comparisons.

Fig 3: RAB10 overexpression shifts the lysosomal population towards anterograde motility.(A-B) Quantification and example kymographs showing the fraction of LC3 moving retrograde, anterograde, or exhibiting bidirectional/stationary motion in the presence of EGFP alone (Control) or EGFP-RAB10. n = 14 neurons; two-way ANOVA with Sidak’s multiple comparisons test (anterograde, P = 0.8445; stationary/bidirectional, P = 0.7328; retrograde, P = 0.2696). (C) Fraction of the motile LAMP1 puncta (moving = 10µm in either direction during a 2 min video) moving anterograde. n = 14 neurons; unpaired t test (P = 0.0297). (D) Within the RAB10-expressing cells, motility of the LAMP1+ puncta either colocalized with RAB10 (+ RAB10) or not (- RAB10). n = 9 neurons; two-way ANOVA with Sidak’s multiple comparisons test (anterograde, P = 0.5595; stationary/bidirectional, P = 0.3011; retrograde, P = 0.9651).

Fig 4: JIP3/4 comigrate with and interact with dynein on autolysosomes.(A) N-terminal region of JIP3 and JIP4, scaled to the primary sequence. Arrows: green, JIP-DIC PLA only; magenta, LC3 only; cyan, LAMP1 only; ochre, LC3 + LAMP1; white, LC3 + LAMP1 + JIP-DIC PLA. (B) Relative enrichment (normalized number of peptides, see Methods for details) for JIP3 (Mapk8ip3) and JIP4 (Spag9) in the proteomics performed by Goldsmith et al., (2022) and Dumrongprechachan et al., (2022). (C-D) Time series demonstrating JIP3 and JIP4 comigration with LC3 (AV marker) and LAMP1 (lysosome marker). (E-F) Example kymographs from the proximal axons of neurons transfected with JIP3 or JIP4. Kymographs depict distance on the x-axis and time on the y-axis. Annotated kymographs (annot.) mirror the above kymographs with the JIP3/4+ puncta paths pseudo-colored for visualization. (G) Quantification of JIP3/4+ puncta moving retrograde (=10µm towards the soma), anterograde (=10µm towards the axon tip), or exhibiting bidirectional/stationary motility (moving <10µm). n = 10 neurons; two-way ANOVA with Sidak’s multiple comparisons test (anterograde, P = 0.8976; stationary/bidirectional, P = 0.0767; retrograde, P = 0.2792). (H) Fraction of motile events (= 10 µm either direction) moving retrograde. n = 10 neurons; unpaired t test (P = 0.9776). (I) Schematic illustrating the proximity ligation assay (PLA). (J-M) Example micrographs and quantifications showing colocalization between LC3, LAMP1, and JIP3-DIC (J-K) or JIP4-DIC (L-M) puncta. n = 20 neurons; one-way ANOVA with Tukey’s multiple comparisons test; JIP3 (LC3 v. LC3 + LAMP1, P < 0.0001; LC3 v. LAMP1, P = 0.7236; LAMP1 v. LC3 + LAMP1, P < 0.0001); JIP4 (LC3 v. LC3 + LAMP1, P < 0.0001; LC3 v. LAMP1, P = 0.9879; LAMP1 v. LC3 + LAMP1, P < 0.0001).

Fig 5: ALS-associated mutations in p62 perturb lysophagy(A) Schematic of p62 domain architecture, highlighting specific ALS-associated mutations in p62 (L341V, P392L, and G425R).(B) Representative images of a-LAMP1 in HeLa that were transiently transfected with p62 siRNA (#1) alongside either WT-p62, L341V, P392L, G425R, or mCherry vector after addition of 750 µM LLOMe for 2 h. Cell outlines were traced. SBs: 10 µm.(C) Quantification of a-LAMP1 area overlapping with p62 area across different conditions (one-way ANOVA, p(WT/L341V) < 0.01, p(WT/P392L) < 0.001, p(WT/G425R) < 0.001, p(WT/mCherry) < 0.0001, n = 3 experiments. Number of cells analyzed per condition = 22).(D) Representative images of a-LC3B in HeLa cells that were transiently transfected with p62 siRNA (#1) alongside FLAG-LAMTOR2 and either WT-p62, P392L, G425R, or mCherry after addition of 750 µM LLOMe for 2 h. Merged channel shows LAMTOR2 and LC3B only. Cell outlines were traced. SBs: 10 µm.(E) Quantification of normalized a-LC3B intensity at LAMTOR2 puncta across different conditions (Kruskal-Wallis test, p(WT/P392L) < 0.01, p(WT/G425R) < 0.05, p(WT/mCherry) < 0.0001, n = 49 cells per condition.(F) A model for p62-dependent lysophagy. p62 is recruited to damaged lysosomes through ubiquitin. Accumulation of p62 on damaged lysosomes results in phase separation. This phase separation of p62 is critically regulated by the small heat shock protein HSP27. HSP27 maintains the liquidity of p62 condensates, facilitating engulfment by autophagosomes.All error bars reflect mean ± SEM. See also Figure S7.