Structural basis of the ATG13–FIP200 interaction.

(A) Structure of the ULK1– ATG13–FIP200 core complex predicted by AlphaFold2. Flexible loop regions in FIP200 were removed from the Figure for clarity. N and C indicate N-and C-terminal regions, respectively. (B) Close-up view of the interactions between ATG13 and FIP200. The bottom panels represent the surface model of FIP200 with the coloring based on the electrostatic potentials (blue and red indicate positive and negative potentials, respectively). (C) ITC results obtained by titration of MBP-ATG13 (363–517 aa) WT or FIP3A mutant into an FIP200 (1–634 aa) solution. (D) Effect of the ATG13 FIP3A mutation on the FIP200 interaction in vivo. atg13 KO HeLa cells stably expressing FLAG-tagged ATG13 WT or FIP3A were immunoprecipitated with an anti-FLAG antibody and detected with anti-FIP200, anti-ULK1, and anti-FLAG antibodies. (E) Relative amounts of precipitated FIP200 in (D) were calculated. Solid bars indicate the means, and dots indicate the data from three independent experiments. Differences were statistically analyzed using Tukey’s multiple comparisons test.

Structural basis of the ULK1–ATG13 interaction.

(A) Structure of the ULK1– ATG13 moiety of the ULK1–ATG13–FIP200 core complex in Fig. 1 A (left) and crystal structure of the yeast Atg1–Atg13 complex (right, PDB 4P1N). (B) Close-up view of the interactions between ATG13MIM(N) and ULK1MIT2 (left) and between ATG13MIM(C) and ULK1MIT1 (right). (C) ITC results obtained by titration of MBP-ULK1 (636–1050 aa) into a solution of WT or ULK2A mutant of MBP-ATG13 (363–517 aa). Due to weak binding, the KD value for the ULK2A mutant was not accurately determined. (D) Effect of the ATG13-FIP3A mutation on endogenous ULK1 levels in vivo. WT or atg13 KO HeLa cells stably expressing FLAG-tagged ATG13 WT or ULK2A mutant were lysed, and indicated proteins were detected by immunoblotting using anti-FIP200, anti-ULK1, and anti-FLAG antibodies. (E) Relative amounts of ULK1 in (D) were normalized with β-actin and calculated. Solid bars indicate the means, and dots indicate the data from three independent experiments. Differences were statistically analyzed using Tukey’s multiple comparisons test.

Structural basis of the ULK1–FIP200 interaction.

(A) Structure of the ULK1– FIP200 moiety of the ULK1–ATG13–FIP200 core complex in Fig. 1 A. The right panel represents the surface model of FIP200 with coloring based on the electrostatic potentials (blue and red indicate positive and negative potentials, respectively). Dotted squares indicate the regions displayed in (B). (B) Close-up view of the interactions between ULK1MIT1 and FIP200 (left) and between ULK1MIT2 and FIP200 (right). In vitro pulldown assay between GST-ULK1 (636–1050 aa) WT or FIP2A mutant with MBP-FIP200 (1–634 aa). (D) Relative amounts of precipitated MBP-FIP200 in (C) were calculated. Solid bars indicate the means, and dots indicate the data from three independent experiments. Differences were statistically analyzed using Tukey’s multiple comparisons test. (E) Effect of the ULK1 FIP2A mutation on the FIP200 interaction in vivo. Ulk1,2 DKO MEFs stably expressing FLAG-tagged ULK1 WT or FIP2A mutant were immunoprecipitated with an anti-FLAG antibody and detected with anti-FIP200, anti-ATG13, and anti-FLAG antibodies. (F) Relative amounts of precipitated FIP200 (left) and ATG13 (right) in (E) were calculated. Solid bars indicate the means, and dots indicate the data from three independent experiments. Differences were statistically analyzed using Tukey’s multiple comparisons test. (G) Halo-LC3 processing assay of ULK1 FIP2A-expressing cells. Ulk1,2 DKO MEFs stably expressing Halo-LC3 and FLAG-tagged ULK1 WT or FIP2A mutant were labeled for 15 min with 100 nm TMR-conjugated Halo ligand and incubated in starvation medium for 1 h. Cell lysates were subjected to in-gel fluorescence detection. (H) Halo processing rate in (G). The band intensity of processed Halo and Halo-LC3 in each cell line was quantified, and the relative cleavage rate was calculated as FLAG-ULK1 WT-expressing cells as 1. Solid bars indicate the means, and dots indicate the data from three independent experiments. Data were statistically analyzed using Tukey’s multiple comparisons test. (I) Colocalization of FLAG-ULK1 WT or FIP2A mutant with FIP200. Ulk1,2 DKO MEFs stably expressing FLAG-tagged ULK1 WT or FIP2A mutant were immunostained with anti-FLAG and anti-FIP200 antibodies. Scale bar, 10 μm.

ATG13–ULK1 and ATG13–FIP200 interactions complement each other in autophagy function.

(A) Schematic representation of the CRISPR-Cas9-mediated KI strategy of ATG13 mutations with FLAG tag. The C-terminally FLAG-tagged coding sequence after the exon 14 of ATG13 with or without FIP3A, ULK2A, or FU5A mutations were knocked in exon 14 of the Homo sapiens ATG13 locus. As the KI cassette expresses NeoR under the hPGK1 promoter, clones that were successfully knocked in were selected by G418. Cas9-gRNA-targeted sites in the exon 14 of H. sapiens ATG13 locus are displayed in dark blue. The homology arm for KI is presented in magenta, and the ATG13 CDS and mutations in red and cyan, respectively. NeoR is displayed in brown. Scale bar, 0.5 kilobase pair (kb). (B) Immunoblot of ATG13-FLAG KI cell lines. WT, atg13 KO, and indicated KI HeLa cells were lysed, and indicated proteins were detected by immunoblotting using anti-FIP200, anti-ULK1, and anti-FLAG antibodies. (C) Colocalization of endogenous levels of ATG13-FLAG mutants with FIP200. Indicated KI cell lines were cultured in the starvation medium for 1 h and immunostained with anti-FLAG and anti-FIP200 antibodies. Scale bar, 10 μm. (D) Halo-LC3 processing assay of ATG13-FLAG KI cell lines. Indicated KI cell lines were labeled for 15 min with 100 nm TMR-conjugated Halo ligand and incubated in starvation medium for 1 h. Cell lysates were subjected to in-gel fluorescence detection. (E) Halo processing rate in (D). The band intensity of processed Halo and Halo-LC3 in each cell line was quantified, and the relative cleavage rate was calculated as FLAG-ULK1 WT-expressing cells as 1. Solid bars indicate the means, and dots indicate the data from three independent experiments. Data were statistically analyzed using Tukey’s multiple comparisons test. (F) Schematic depiction 0of the difference between the mammalian ULK complex and the yeast Atg1 complex. Mammalian ATG13 binds to two FIP200s within the same FIP200 dimer, contributing to the stability of one ULK1 complex. Conversely, budding yeast Atg13 binds to two Atg17s within a different Atg17 dimer, allowing for endlessly repeated Atg13-Atg17 interactions. ATG101 in the ULK1 complex and Atg31-29 in the Atg1 complex are omitted for simplicity. ATG13/Atg13 is shown in yellow, ULK1/Atg1 in magenta, and FIP200/Atg17 in green. Black lines represent interactions.