Fig 2: Western blot analysis of anti-LC3 antibody specificity. Each framed box represents three membrane pieces (LC3A/LC3B/LC3C) that were developed with the indicated antibody (right margin) simultaneously. Membrane pieces for each LC3 had been cut from the same master membrane loaded several times with only one of the LC3s to assure equal transfer efficiency. The loaded amount per lane and the expected size (in kDa) are indicated. The employed antibodies were: anti-LC3A from Cell Signaling (#4599), anti-LC3B from Sigma (#L7543), anti-LC3C from Abcam (#ab150367).

Fig 3: Analysis of the cellular localization of LC3C and LC3B in response to pharmacological sirtuin inhibition. Primary human fibroblasts were treated with BafA1, Ex527 or TSA and analyzed as in Figure 3. (A) Cellular localization of LC3C, LC3B and p62 after treatment with BafA1 alone. (B) Cellular localization of LC3C, LC3B and p62 after treatment with Ex527 and BafA1. (C) Cellular localization of LC3C, LC3B and p62 after treatment with TSA and BafA1. (D) Quantification of the relative nuclear fraction of LC3C and LC3B in cells treated as indicated. Significant changes (by two-way ANOVA) versus the control: ### p = 0.001; significant changes between the LC3s: ** p = 0.01; *** p = 0.001. (E) Relative co-localization of LC3C and LC3B with p62 quantified by image analysis. Statistical analysis was done as in D. Data in (D) and (E) were derived from three photographed, independent experiments (n = 3), in which 12–45 cells per image were analyzed. (F) Magnified images of cells immunostained for LC3C or LC3B including the chromatin counterstain (DAPI).

Fig 4: In silico sequence and binding partner analysis for the ATG8 protein family. (A) Phylogenetic tree of the ATG8 gene family featuring H. sapiens LC3A, LC3B, LC3B2, LC3C, GABARAP, GABARAP-L1, and GABARAP-L2, D. melanogaster ATG8A and ATG8B as well as S. cerevisiae ATG8. (B) Protein sequence alignment indicating identical (*), conserved (:) and semi-conserved (.) amino acids. Residues that have been found essential for ATG7 binding are indicated by a purple arrow, the conserved lipidation site glycine is marked with blue. (C) Meta-analysis of the amino acid composition of all putative (4-amino-acid) LIR motifs of the indicated number of recognized, exclusive ATG8 interactors as published [18]. The relative amino acid abundance for each position of a LIR motif was calculated and is shown by the indicated color-coding.

Fig 5: Analysis of the cellular localization of LC3A, LC3B, LC3C after siRNA-mediated LC3A knock-down. Primary human fibroblasts were treated with BafA1 after transfection with (A) scrLC3A-RNA and (B) siLC3A-RNA, (C) siLC3A-RNA and Ex527 treatment, (D) siLC3A-RNA and starvation, and (E) siLC3A-RNA and starvation and Ex527 treatment. (F) Quantification of the relative nuclear fraction of LC3B and LC3C in cells treated as indicated. (G) Relative co-localization of LC3B and LC3C with p62 quantified by image analysis. Significant changes (by two-way ANOVA) versus scrLC3A: ** p ≤ 0.01; *** p ≤ 0.001; significant changes versus siLC3A: # p ≤ 0.05; ## p ≤ 0.01. (H) LC3A protein expression in cells transfected with scrLC3A-RNA or siLC3A-RNA. Tubulin was used as loading control. (I) Control of knock-down and antibody specificity under immunocytochemistry conditions. Whole-cell signal intensities were quantified and evaluated by two-way ANOVA. Data in (F), (G) and (I) were derived from three photographed, independent experiments (n = 3), in which 12–45 cells per image were analyzed.