Fig 2: Atg9 dynamics is impaired in CAPNS1-depleted cells. (A) ShCAPNS1 and control cells were co-transfected with GFP-Atg9 and HcRed-LC3. 24 h later, the cells were analysed using a confocal microscope. Images were acquired every 2 min over a 50 min time interval. The first (T0) and last (T90) merged pictures of the 2 h time-lapse experiments are shown. Scale bars: 20 µm. Magnifications of boxed areas are shown in the right panels. Arrows indicate ATG9 vesicles adjacent to or fused with LC3 vesicles. (B) Quantification of GFP-Atg9 and HcRed-LC3 colocalization by Pearson's coefficient determination in control and CAPNS1-depleted U2OS cells, before and after thapsigargin treatment. (C) Images sequences (T21, T23, T25) of the GFP-ATG9/HcRed-LC3 time-lapse experiment made using control and CAPNS1-depleted U2OS cells. The arrow indicates one ATG9 vesicle that meets and fuses with a LC3-positive vesicle in control cells. In shCAPNS1 cells, ATG9- and LC3-positive vesicles are more stationary. (D) U2OS cells stably expressing Flag-Atg9 were transfected with a CAPNS1-specific siRNA or a control siRNA. 48 h later, the cells were treated for 1 h with DMSO or 100 nM thapsigargin before lysis and immunoprecipitation with an anti-Flag antibody. The immunoprecipitation products were analysed by western blot to visualize Flag-Atg9, endogenous TfR and Vps34.

Fig 3: Atg9/Bif-1 trafficking from the Golgi is impaired in CAPNS1-depleted cells. (A) Control and shCAPNS1 U2OS cells were transiently co-transfected with AmCyan-Bif-1 and HcRed-LC3. 16 h later, the cells were fixed and analysed by a confocal microscope. Arrows indicate areas of colocalization of AmCyan-Bif-1 and HcRed-LC3. Scale bars: 20 µm. (B) 100 cells for each sample were considered and the number of cells where Bif-1 forms aggregates or single dots was counted. The graph represents the means and s.d. of three independent experiments. (C) U2OS cells stably expressing Flag-Atg9 were transiently silenced with control or CAPNS1-specific siRNA and then transfected with AmCyan-Bif-1. 24 h later the cells were treated for 1 h with or without 100 nM thapsigargin, and then fixed and analysed by immunofluorescence to visualize endogenous GM130, Flag-Atg9 and AmCyan-Bif1. Representative confocal merged images are shown. Scale bars: 20 µm. (D) 50 cells for each sample were considered and the number of cells where Atg9/Bif-1 positive vesicles colocalize with endogenous GM130 was counted. Mean and s.d. of three independent experiments.

Fig 4: Early endosome dynamics is impaired in CAPNS1-depleted cells. (A) Control and shCAPNS1 U2OS cells were grown on Petri dishes, treated with commercial baculovirus reagent expressing GFP-Rab5 protein and RFP-KDEL. 24 h later, the cells were analysed using a confocal microscope to detect the dynamics of Rab5-positive vesicles. Images were acquired every 2 min over a period of 50 min. The first (T0) and last (T50) pictures of each-time lapse experiment are shown. Scale bars: 20 µm. (B) Corrected total cell fluorescence (CTCF) levels of GFP-Rab5 measured on 100 fixed control and shCAPNS1 cells. The graph indicates the percentage of cells that show a CTFC =3×106; the error bars represent s.d. of three different independent experiments; P<0.01.

Fig 5: CAPNS1 depletion perturbs Golgi stack distribution. (A) Control, shCAPNS1 and CAPNS1-rescued U2OS cells were fixed and stained with GM130 antibody. Hoechst dye was used to stain nuclei. A fluorescence microscope was used to acquire pictures; Scale bars: 20 µm. Lower and right panels are magnifications of boxed areas. 100 cells for each sample were considered and the percentage of cells with Golgi stacks surrounding the nucleus was quantified. (B) Means and s.d. of three independent experiments.