Fig 2: ATG5 overexpression neuroprotects and modulates motor and secretory pathway-related proteins and cytoskeleton.a Bar graphs of the mean average percentage of survival of NSC34 cells, transfected with CMV-GFP or CMV-ATG5 plasmids, at 18 h after nocodazole (N) treatment and analyzed by MTT assay (n = 4; *p < 0.001 vs. control, one-way ANOVA). b Analysis of processed CTSD by western blot in transfected cells treated with vehicle (Veh) or nocodazole. c Western blot showing the levels of different autophagy hallmarks at the ipsilateral side of the spinal cord ventral horn from non-injured animals injected with AAVrh10-GFP or AAVrh10-ATG5 at 3 weeks post injection. Note LAMP1, ATG5-ATG12, ATG5, LC3II abundance, and normal p62 levels. d Western blot and corresponding bar graphs of the quantification of different proteins related to autophagy (p62, LAMP1, LC3, Beclin1), cytoskeleton (DCTN1, KIF5c), the secretory pathway (Sec31, EEA1), and lysosome function (CTSD) in the spinal cord from RA-injured animals injected with AAVrh10-GFP or AAVrh10-ATG5. e Left, Microphotographs showing green fluorescent Nissl stained MNs at the contralateral side and the ipsilateral side of the spinal cord ventral horn from RA-injured animals injected with AAVrh10-GFP or AAVrh10-ATG5. Right, Bar graphs showing the percentage of MN survival at the ipsilateral side with respect to the contralateral side of the same sample. f Representative microphotographs of MNs stained with a-tubulin (a-tub) or acetyl-a-tubulin (Ac-a-tub) from the different groups and associated bar graphs of the ratio of the percentage of immunoreactivity for Ac-a-tub with respect to a-tub. Scale bar = 10 µm

Fig 3: Alterations in protein trafficking in the in vitro model.a Immunoblots and histogram showing the levels of EEA1, Sec31A, and p115 in control NSC34 cells and in cells treated with nocodazole (N), rapamycin (R), or the sequential combination (N2R3). b Microphotographs showing LAMP1 (green) and CTSB (red) immunostaining and co-localization (merged images) in indicated conditions. Below, graph bars of immunofluorescent quatification for each marker. Scale bar = 15 µm. c Representative confocal images of LAMP1 (red) co-immunolocalization (merged) with EEA1 (magenta) and giantin (green). Insets are two-fold magnifications of regions in boxes. Note that LAMP1 and giantin co-localized in cells treated with nocodazole. Scale bar = 30 µm

Fig 4: Impaired autophagy flux and aberrant lysosomal protein glycosylation after RA.a Schematic representation of the autophagy flux (modified from ref. 13). Expression of the mCherry-GFP-LC3 reporter gene will generate yellow puncta upon activation of autophagy at the autophagosome/amphisome and red puncta in the autolysosome if the autophagy flux continues correctly and the lysosomes achieve their usual acidic pH which quenches the green GFP fluorescence. b Microphotographs of representative MNs from sham-operated control animals treated with rapamycin (Rapa) for 3 days before sacrifice, from RA injured and control sham rats at 5 dpi; all rats were infected with AAVrh10-mCherry-GFP-LC3. c Left, Bar graph of manual quantification of yellow and red puncta within MNs at L4 spinal cord segment in both conditions. Right, Bar graph of autophagic flux measured as the ratio of red to yellow puncta per MN. Non-parametric t-test was used to compare the statistical significance between the control and avulsed animal. d Left, Immunoblot for glycosylated and unglycosylated v-ATPase, LAMP1, and p62. Note the slight difference in electrophoretic mobility of higher LAMP1 band between control and RA samples and the different general band pattern. Right, Bar graph showing the average fold change protein level ± SEM in control (C) and root avulsed (RA) samples normalized to actin levels (n = 4). Scale bar = 10 µm; zoom = 5 µm; *p < 0.05 vs. control, ***p < 0.001 vs. control (Student’s t-test)