Fig 2: C9orf72 induces autophagy via the ULK1 complex HEK293 cells treated with non‐targeting (Ctrl) or FIP200 siRNA were co‐transfected with EGFP‐LC3 and either empty vector control (EV), Myc‐C9orf72S, or Myc‐C9orf72L. 24 h post‐transfection cells were treated with vehicle or 100 nM BafA1 for 6 h. Samples were lysed and subjected to SDS–PAGE and immunoblot. Autophagy levels were determined by immunoblot for EGFP‐LC3‐I and II. Expression of Myc‐C9orf72 was confirmed using anti‐Myc (* indicates a nonspecific band). FIP200 knockdown was confirmed using anti‐FIP200 antibodies. α‐tubulin was used as a loading control.HeLa cells treated with non‐targeting (Ctrl) or FIP200 siRNA were co‐transfected with empty vector (EV), Myc‐C9orf72S or Myc‐C9orf72L (red), and EGFP‐LC3 (green) to label autophagosomes. As positive control, EV‐transfected cells were treated for 3 h with Torin1 (250 nM). Autophagy was quantified as the number of EGFP‐LC3‐positive autophagosomes per cell from 3 independent experiments (mean ± SEM; one‐way ANOVA with Fisher's LSD test: ****P ≤ 0.0001, N (cells) = Ctrl/EV: 73; FIP200/EV: 76; Ctrl/EV/Torin1: 73, FIP200/EV/Torin1: 70; Ctrl/C9orf72L: 71; FIP200/C9orf72L: 74; Ctrl/C9orf72S: 72; FIP200/C9orf72S: 72). Scale bar = 20 μm. FIP200 knockdown was confirmed by immunoblot (Appendix Fig S2). Source data are available online for this figure.

Fig 3: C9orf72 interacts with the ULK1 initiation complex A–CCell lysates of HEK293 cells co-transfected with FLAG-FIP200 (A), or HA-ULK1 (B) and either empty vector control, Myc-C9orf72S, or Myc-C9orf72L or with Myc-ATG13 (C) and either empty vector control, FLAG-C9orf72S, or FLAG-C9orf72L were subjected to immunoprecipitation with anti-Myc (A and B) or anti-FLAG (C) antibodies. Immune pellets were probed for Myc-C9orf72 (A and B), FLAG-C9orf72 (C), FLAG-FIP200 (A), HA-ULK1 (B), or Myc-ATG13 (C) on immunoblots.D, ECell lysates of HEK293 cells transfected with empty vector control, Myc-C9orf72S, or Myc-C9orf72L were subjected to immunoprecipitation with anti-Myc antibodies. Immune pellets were probed for Myc-C9orf72 and endogenous FIP200, ULK1, and ATG13. There are multiple alternatively spliced forms of ATG13 (Jung et al, 2009; Alers et al, 2011); ATG13* is most likely a smaller alternative spliced form of ATG13 that is enriched by interaction with C9orf72.F–HHeLa cells transfected with HA-ULK1 or Flag-FIP200 and Myc-C9orf72 or with Myc-ATG13 and EGFP-C9orf72 were fixed and processed for PLA analysis. Transfections were laced with mVenus to enable identification of transfected cells for analysis where required (green). PLA signals (red) were counted per cell (mean ± SEM; one-way ANOVA with Fisher's LSD test, ****P = 0.0001; N (cells) = (F) C9orf72S: 22; C9orf72L: 18; EV+FIP200: 18; C9orf72S+FIP200: 18; C9orf72L+FIP200: 17; (G) C9orf72S: 21; C9orf72L: 20; EV+ULK1: 20; C9orf72S+ULK1: 22; C9orf72L+ULK1: 20; (H) C9orf72S: 11; C9orf72L: 10; EV+ATG13: 11; C9orf72S+ATG13: 11; C9orf72L+ATG13: 11). Scale bar = 10 µm; see also Fig EV2.I–K 35S-radiolabeled recombinant FIP200-6xHis segments (I), Myc-ATG13 (J), or HA-ULK1 (K) were added to GST, GST-C9orf72S, and GST-C9orf72L immobilized on glutathione-coated beads. 35S-radiolabeled recombinant proteins were visualized by phosphorimager (top panels). Coomassie-stained GST, GST-C9orf72S, and GST-C9orf72L in the pull-down samples are shown (bottom panels). The identity of the Coomassie protein bands was confirmed by mass spectrometry (# indicates E. coli DnaK chaperonin; * indicates E. coli 60kD chaperonin; Appendix Fig S3). Source data are available online for this figure.

Fig 4: C9orf72 mediates interaction of Rab1a with the ULK1 complex HEK293 cells treated with non-targeting (Ctrl) or C9orf72 siRNA were transfected with HA-ULK1, Myc-Rab1a, or HA-ULK1 + Myc-Rab1a as indicated. Transfections were laced with mVenus to enable identification of transfected cells for analysis (green). Transfected cells were probed with anti-HA and anti-Myc antibodies and processed for PLA. PLA proximity signals (red) per cell were determined from 3 independent experiments (mean ± SEM; one-way ANOVA with Fisher's LSD test, ****P = 0.0001; N (cells) = Ctrl/HA-ULK1: 125, Ctrl/Myc-Rab1a: 149, Ctrl/HA-ULK1 + Myc-Rab1a: 163, C9orf72/HA-ULK1: 136, C9orf72/Myc-Rab1a: 133, C9orf72/HA-ULK1 + Myc-Rab1a: 155). Scale bar = 20 µm.HeLa cells treated with non-targeting (Ctrl) or C9orf72 siRNA were co-transfected with mCherry-FIP200 (red) and empty vector (EV) or Myc-Rab1a(Q70L) (green). Translocation of the ULK1 complex was quantified as the number of mCherry-FIP200-positive puncta per cell from 4 independent experiments (mean ± SEM; one-way ANOVA with Fisher's LSD test, *P = 0.05, ****P = 0.0001; N (cells) = Ctrl/EV: 101; Ctrl/Q70L: 106; C9orf72/Q70L: 42). C9orf72 knockdown was determined by RT–qPCR (Appendix Fig S2). Scale bar = 10 µm.HeLa cells treated with non-targeting (Ctrl) or C9orf72 siRNA were co-transfected with EGFP-LC3 (green) and empty vector (EV) or Myc-Rab1a(Q70L) (red). Autophagosomes were quantified as the number of EGFP-LC3-positive puncta per cell (mean ± SEM; one-way ANOVA with Fisher's LSD test, **P = 0.01, ****P = 0.0001; N (cells) = Ctrl/EV: 44; Ctrl/Q70L: 32; C9orf72/Q70L: 47). C9orf72 knockdown was determined by RT–qPCR (Appendix Fig S2). Scale bar = 10 µm.

Fig 5: C9orf72 regulates translocation of the ULK1 complex via Rab1a A, BHeLa cells (A) or SH-SY5Y neuroblastoma cells (B) treated with non-targeting (Ctrl) or Rab1a siRNA were co-transfected with mCherry-FIP200 (red) and empty vector (EV), Myc-C9orf72L, or Myc-C9orf72S (green). As positive control, EV-transfected cells were treated for 3 h with Torin1 (250 nM). Translocation of the ULK1 complex was quantified as the number of mCherry-FIP200-positive puncta per cell [A, HeLa, mean ± SEM; one-way ANOVA with Fisher's LSD test; ns, not significant; ****P = 0.0001; N (cells from 3 independent experiments) = Ctrl/EV: 81; Ctrl/EV/Torin1: 48; Ctrl/C9orf72L: 86; Ctrl/C9orf72S: 48; Rab1a/EV: 79; Rab1a/EV/Torin1: 68; Rab1a/C9orf72L: 78; Rab1a/C9orf72S: 52; B, SH-SY5Y: mean ± SEM; one-way ANOVA with Fisher's LSD test; ns, not significant; **P = 0.01; ****P = 0.0001; N (cells from 2 independent experiments) = Ctrl/EV: 70; Ctrl/EV/Torin1: 56; Ctrl/C9orf72L: 45; Ctrl/C9orf72S: 43; Rab1a/EV: 63; Rab1a/EV/Torin1: 55; Rab1a/C9orf72L: 44; Rab1a/C9orf72S: 37]. Rab1a knockdown was confirmed by RT–qPCR (Appendix Fig S2). Scale bar = 10 µm.CSH-SY5Y neuroblastoma cells treated with non-targeting (Ctrl) or Rab1a siRNA were co-transfected with EGFP-LC3 (green) and empty vector (EV), Myc-C9orf72L, or Myc-C9orf72S (red). As positive control, EV-transfected cells were treated for 3 h with Torin1 (250 nM). Autophagosomes were quantified as the number of EGFP-LC3-positive puncta per cell (mean ± SEM; one-way ANOVA with Fisher's LSD test; ns, not significant; *P = 0.05, ***P = 0.001, ****P = 0.0001; N (cells from 2 independent experiments) = Ctrl/EV: 102; Ctrl/EV/Torin1: 92; Ctrl/C9orf72L: 97; Ctrl/C9orf72S: 76; Rab1a/EV: 102; Rab1a/EV/Torin1: 107; Rab1a/C9orf72L: 101; Rab1a/C9orf72S: 87). Rab1a knockdown was confirmed by RT–qPCR (Appendix Fig S2). Scale bar = 10 µm.