Fig 2: ATG7 and ATG9B may play autophagy-independent roles in HNSCC. (a) Correlation of different autophagy-related genes with transcription level based on the TCGA database. (b) The mRNA levels of ATG7 and ATG9B in HNSCC tumor and normal tissues. (c–f) The correlation between clinicopathological features and expression of ATG7 and ATG9B in HNSCC patients. (g) Kaplan-Meier analysis of the mRNA expression of ATG7 and ATG9B on the overall survival rate of HNSCC patients.

Fig 3: 20A and chloroquine act in concert to trigger cell death. (a) HeLa cells (left) and A549 cells (right) were treated with the indicated concentration of 20A with or without 25 µM of chloroquine for 24 h. Cell death was evaluated by scoring the percentage of propidium iodide (PI) positive cells after flow cytometer analysis. The data represents the mean ± SD of 6 values obtained from three independent experiments each performed in duplicate (** p value < 0.01 using Kruskal-Wallis test, with post-hoc Dunn’s analysis). (b) U20S cells were treated with the indicated concentration of 20A with or without 25 µM of chloroquine for 16 h. Left panel, cell death was evaluated by scoring the percentage of propidium iodide positive cells after flow cytometer analysis. The data represents the mean ± SD of 9 values obtained from three independent experiments each performed in triplicate (p value **** < 0.0001, using Mann-Whitney test). Right panel, Cell death was evaluated after 16 h treatment with 20A ± chloroquine by scoring the percentage of TMRMLow/ DAPILow (early apoptosis) and TMRMLow/ DAPIHigh (late apoptosis) cells following flow cytometer analysis. The data represents the mean ± SD of 9 values obtained from three independent experiments each performed in triplicate (p value **** < 0.0001, using Mann-Whitney test). (c) Autophagy-proficient (sgNTC) and -deficient (sgATG5 and sgATG7) HeLa cells were treated with 5 µM 20A either in the presence (25 µM) or absence of chloroquine for 24 h. Left panel, silencing of key autophagy proteins, ATG5 and ATG7 was checked by western blotting. The western blots were quantified and given as ratio of ATG5-ATG12/Actin β or ATG7/Actin β. Right panel, cell death was evaluated by scoring the percentage of propidium iodide positive cells after flow cytometer analysis. The data represents the mean ± SD of nine values obtained from three independent experiments each performed in triplicate (p value * < 0.05 and p value *** < 0.001, ns: not significant using Mann-Whitney test).

Fig 4: PTEN and ATG7 are direct targets of mir‑20a‑5p, and mir-20a-5p regulates PTEN, ATG7, and AKT/mTOR signaling in EPCs induced by high glucose.a, c Schematic representation of the putative target sites of mir-20a-5p in the 3′-UTRs of ATG7 and PTEN, and the sequence of mir-20a-5p mutant (indicated as Mut). b, d Luciferase assay of cells cotransfected with wild-type (WT) or mutated (Mut) ATG7 3′-UTR or PTEN 3′-UTR reporter construct and mir-20a-5p or the control for 48 h. e Western blot analysis of PTEN and ATG7 protein levels in EPCs treated with mir-20a-5p inhibitor or mimics. f Western blot analysis of PTEN, ATG7, p-AKT, and p-mTOR protein levels in EPCs treated with mir-20a-5p inhibitor or mimics under high-glucose condition. Inh inhibitor, Mic mimic, *P < 0.05, **P < 0.01, n = 3.

Fig 5: TAX1BP1 is required for NBR1 flux in lipidation-deficient cells (related to Figs 3 and 4) K562-derived extracts prepared from wild-type (WT) and clonal deletion isolates were resolved by SDS–PAGE followed by immunoblotting (IB) with indicated antibodies. All samples were normalized by total protein using a BCA assay prior to loading. ATG7KO/TAX1BP1KO cells expressing tf-NBR1 were transfected with TagBFP-V5-TAX1BP1 and analyzed for red:green ratio and BFP expression. Plotted are the median red:green ratios of BFP-negative and BFP-positive cells relative to ATG7KO cells alone. Bar graphs represent mean ± SD from three independently generated ATG7KO/TAX1BP1KO cell lines. P values were determined using a ratio paired t-test. **P < 0.01; ns, not significant. n = 10,000 cells per sample. TAX1BP1KO and ATG7KO/TAX1BP1KO K562 cells expressing tf-NBR1 were nucleofected with TagBFP, TagBFP-ATG7, or TagBFP-ATG7C572S and analyzed for red:green ratio and BFP expression at 8 h post-nucleofection. Median values for each sample are identified by a black line within each violin. The red dotted line corresponds to the red:green ratio of TAX1BP1KO cells expressing TagBFP. The solid red line corresponds to the red:green ratio of ATG7KO/TAX1BP1KO cells expressing TagBFP. n > 5,000 cells for each sample.Wild-type K562 cells expressing tf-LC3 or tf-NDP52 were transfected with TagBFP or TagBFP-RavZ. BFP-positive cells were analyzed for red:green ratio. The median red:green ratio of each sample was used to calculate flux relative to WT + TagBFP. Bar graphs represent mean ± SD from three independent experiments. n > 10,000 cells per sample. P values were determined using a ratio paired t-test. *P < 0.05; **P < 0.005; ns, not significant.Representative confocal micrographs (as maximum intensity projections) of indicated K562 genotypes expressing tf-NBR1. Selected regions (white box) of micrographs are shown as single and merged channels from fluorescence microscopy. LC3, magenta; NBR1, green; Hoechst, blue. Scale bars: large panels, 5 µm; small panels, 1 µm.K562-derived extracts prepared from wild-type (WT) and clonal deletion isolates were resolved by SDS–PAGE followed by immunoblotting (IB) with indicated antibodies. All samples were normalized by total protein using a BCA assay prior to loading.Total RNA was extracted from the indicated cell lines. Equal amounts of subsequently generated cDNA were analyzed for NBR1 expression levels. All values were normalized to GAPDH expression levels using the ??C method. RPL0 was included as an unchanged control. Bar graphs represent mean ± SD. n = 3 technical replicates.HEK293T cells with indicated genotypes were treated with Bafilomycin A1 (BafA1) or DMSO for 18h. Extracts prepared from wild-type (WT) and deletion cell lines were resolved by SDS–PAGE followed by immunoblotting (IB) with indicated antibodies. All samples were normalized by total protein using a BCA assay prior to loading.Quantitation of protease protection data for tubulin from experiments in Fig 4E. Bar graphs show the mean ± SD of each sample from four independent experiments. Source data are available online for this figure.