Fig 2: Binding of ATG16L1 to lipids is required for PAS recruitment ATG16L1−/− stably expressing the indicated Flag‐S‐ATG16L1 constructs were amino acid starved for 2 h prior to immunofluorescence analyses using antibodies against Flag tag (green) and ATG5 (red). Scale bar: 9 μm. Right panels show quantifications of average number of Flag‐ and ATG5‐positive dots per cell.Cells, as in (A), were immunostained using antibodies against S tag (to detect ATG16L1, red) and WIPI2 (green). Scale bar: 9 μm. Lower panel shows quantification of average number of WIPI2‐positive dots per cell.Average intensities of individual Flag‐ATG16L1 dots in (A). Underlying grey circles represent individual data points. Dots were quantified (n = 1,225 for ATG16L1WT; n = 334 for ATG16L1LD).WIPI2−/− cells reconstituted with the indicated Flag‐S‐ATG16L1 constructs were amino acid starved for 2 h prior to immunofluorescence analyses using antibodies against Flag tag (green) and FIP200 (red). Scale bar: 9 μm. Lower panel shows quantification of average number of Flag‐positive dots per cell.Data information: Quantifications depict means and error bars (SEM) from at least three independent experiments. *P ≤ 0.05, ***P ≤ 0.001, ****P ≤ 0.0001 (pairwise unpaired Student's t‐test).

Fig 3: PAS targeting activity of ATG16L1 lies within its CCD Schematic presentation of ATG16L1 fragments and mutants used in this study. The following fragments encompassed the indicated amino acids: wild type (WT): 1–623; ?1: 39–632; ?2: 120–623; ?3: 206–623 and ?4: 336–623. Mutant ?2?182–205 consists of the ?2 fragment with an additional deletion in amino acids 182–205. All constructs contained a Flag-S tag at the N-terminal end.Fragments depicted in (A) were expressed in U2OS cells and amino acid starved for 5 h followed by fixation and immunostaining against S tag to detect S-ATG16L1. Scale bar: 10 µm.U2OS cells expressing Flag-ATG16L1?2 or Flag-ATG16L1?2?182–205 were treated as in (B) and stained using antibodies against Flag tag (to detect ATG16L1, green) and FIP200 (red). Scale bar: 10 µm.Protein–protein interaction assay in 293T cells transiently transfected with the indicated Flag-S-tagged ATG16L1 constructs. S tag pull-down was performed and protein complexes were analysed by immunoblotting using the indicated antibodies.Homodimerisation assay in 293T cells transiently transfected with ATG16L1-GFP and the indicated Flag-S-tagged ATG16L1 constructs. S tag pull-down was performed and protein complexes were analysed by immunoblotting using the indicated antibodies.Homodimerisation assay similar to (E).Flag-ATG16L1WT or Flag-ATG16L1?182–205 were stably expressed in ATG5-/- cells and analysed by immunofluorescence using antibodies against Flag tag to detect ATG16L1. Scale bar: 9 µm.

Fig 4: ATG16L1 binds liposomes through CCD sequences Sequence alignment of ATG16L1 CCD segment from various species and ATG16L2 (Sc: Saccharomyces cerevisiae; Xt: Xenopus (Silurana) tropicalis; Hs: Homo sapiens; Mm: Mus musculus). Residues that mediate WIPI2b and FIP200 binding are highlighted in orange and magenta, respectively. Cyan-shaded residues (I171, K179 and R193) are exposed conserved residues predicted to not contribute to the dimer-dimer interface and are mutated in this study. Cherry-shaded residues are mutated in Figs 6 and 7.Microscopy-based protein-liposome binding assay. ATG16L1-GFP immobilised on beads incubated in the presence of rhodamine-labelled liposome preparations containing the indicated phosphoinositides. Scale bar: 50 µm.Quantification of relative liposome binding in (B).Structural modelling of ATG16L1 residues 160–205 (magenta helix) in the presence of lipid bilayer (green lines). Highlighted residues include I171, K179 and R193 as sticks, which are mutated in this study. A PI3P molecule is shown as a yellow stick embedded in the lipid bilayer and interacting with the highlighted positively charged residues of ATG16L1.Microscopy-based protein–liposome binding assay as in (B). ATG16L1WT- and ATG16L1LD-GFP immobilised on beads were incubated with rhodamine-labelled, PI3P-positive liposome preparations. Scale bar: 50 µm. Right panel shows quantification of liposome binding relative to ATG16L1WT from three independent experiments including SEM values.Protein–protein interaction assay in 293T cells transiently transfected with the indicated S-tagged ATG16L1 constructs. S tag pull-down was performed and protein complexes were analysed by immunoblotting using the indicated antibodies.Dimerisation assay in 293T cells transiently transfected with ATG16L1-GFP and the indicated S-tagged ATG16L1 constructs and analysed as in (F).Data information: Quantifications depict means and error bars (SEM) from at least three independent experiments. *P = 0.05, **P = 0.01, ***P = 0.001 (pairwise unpaired Student's t-test).

Fig 5: FIP200 is a target of miR-224-3p.(A) Predicted binding sequences for miR224-3p in the FIP200 3'-UTR. (B) Dual-luciferase reporter vectors containing the wild-type or mutated 3'-UTR fragments of FIP200 cloned into the pmirGLO-luciferase plasmids. Luciferase activity was assayed 48 h after co-transfection with either wild-type (WT) or mutated (MUT) 3'-UTR containing plasmids and miR224m or NCm in HeLa cells. Data is represented as mean ± SD of n = 3 independent experiments. miR-224m = miR-224-3p mimetic; NCm = miRNA mimicsnegative control. *p < 0.05 as determined using student’s 2-tailed t-test.