Atg8 regulates the level of autophagy. (A) Strains A8-1, A8-2, and A8-3 showed intermediate Atg8 protein levels. Yeast cells were grown in rich medium to midlog phase and then starved in nitrogen-starvation medium for 4 h. Protein extracts were analyzed by immunoblotting with anti-Atg8 antiserum. Atg8–PE, Atg8 conjugated to PE. (B) Atg8 limits the level of autophagy. Yeast cells were grown in rich medium to midlog phase and then shifted to nitrogen-starvation medium. At the indicated time points, samples were collected and tested by the Pho8Δ60 assay. Specific units, specific activity units (μmoles phosphate/mg/min) normalized to protein concentration. Error bar, SEM from six independent repeats. (C) The Cvt pathway is normal in strains A8-1, A8-2, and A8-3. Yeast cells grown in rich medium were collected at midlog phase; protein extracts were analyzed by immunoblotting with anti-Ape1 antiserum. The locations of precursor and mature Ape1 are indicated.

Atg8 controls the size of the autophagosome. (A and B) Lower amounts of Atg8 reduce the sizes of autophagosomes. Yeast cells were grown in rich medium to midlog phase and then starved in nitrogen-starvation medium for 4 h, fixed in potassium permanganate, and processed for EM. (A) Representative EM images from pep4Δ strains. Autophagic bodies accumulated in A8-1, A8-2, A8-3, and wild-type background strains. No autophagic bodies were found in atg8Δ background cells. Scale bar, 1 μm. (B) Quantification of autophagic body size. The average radii of cross-sections of autophagic bodies are shown; Error bar, SEM; n > 200. Autophagic bodies in the A8-1, A8-2, and A8-3 strains were significantly smaller than those of the wild-type (WT) background. (C) Atg8 level does not limit the number of autophagosomes produced. Yeast cells starved for 2 and 4 h were collected and processed for EM. The average numbers of autophagic bodies per vacuole section are shown. Error bar, SEM; n > 100. Similar numbers of autophagic bodies were observed in strains with different amounts of Atg8.

Dynamics of GFP-Atg8 during autophagosome formation. Yeast cells were grown in rich medium to midlog phase, then starved for 1 h, immobilized on concanavalin A-treated cover slips, and incubated in starvation medium on a depression (concave) slide. Image stacks were collected at the indicated time points; only the images with GFP-Atg8 puncta (if present) in focus are shown. (A) The emergence and disappearance of GFP-Atg8 puncta correspond to autophagosome formation. GFP-Atg8 and mCherry-prApe1 were expressed under their endogenous promoters in wild-type cells. GFP-Atg8 was recruited to the PAS, marked by the presence of the cargo mCherry-prApe1; after some GFP-Atg8 was released, the cargo was delivered to the vacuole and disassembled. The white arrow indicates the GFP-Atg8 punctum being tracked. (B) The dynamic pattern is absent in atg1Δ cells. GFP-Atg8 and mCherry-prApe1 were expressed in atg1Δ cells. GFP-Atg8 and mCherry-prApe1 signals persisted without a significant decrease in fluorescence in their punctate structures during 30 min of observation. DIC, differential interference contrast. Scale bar, 1 μm.

The majority of GFP-Atg8 at the PAS is released during autophagosome formation. (A) The decrease of signal intensity of GFP-Atg8 puncta is caused by the release of GFP-Atg8. GFP-Atg8 was expressed in atg11Δ pep4Δ cells. Yeast cells grown to midlog phase in nutrient-rich medium were collected and starved in nitrogen-starvation medium for 30 min. Microscopy observation was performed as in Figure 3. The signal decreases were evident in atg11Δ pep4Δ cells, which are defective in degradation of autophagic bodies. (B and C) The significant decrease in fluorescence is not an artifact due to photobleaching. The intensities of GFP-Atg8 puncta in live cells and fixed cells were quantified. For live cells, the peak intensities were normalized as 100%, and the peaks were aligned to time 0. For fixed cells, the initial intensities were normalized as 100%. In fixed cells, photobleaching caused less than a 10% reduction in fluorescence after 5 min. In atg11Δ pep4Δ cells (B) and in strain GA8 (GFP-Atg8 in atg8Δ; C), the remaining fluorescence at the PAS was clearly below 50% at 5 min after peak. (D) Atg8 deconjugation is required for the release of Atg8 and completion of autophagosomes. GFP-Atg8ΔR and mCherry-prApe1 were expressed in atg4Δ atg8Δ cells. Yeast cells were starved for 1 h. Microscopy observation was performed as in Figure 3. GFP-Atg8 and mCherry-prApe1 persisted in a punctate structure during 20 min of observation. Scale bar, 1 μm.

Amount of Atg8 at the PAS controls the level of autophagy. (A) Different amounts of GFP-Atg8 were expressed in strain A81-GA8 and GA8. GFP-Atg8 was expressed under the promoter used in strain A8-1 (strain A81-GA8) or its own promoter (strain GA8) in atg8Δ cells. Yeast cells were collected at the indicated time points after starvation. Protein extracts were analyzed by immunoblotting with anti-GFP antibody. GFP refers to the GFP moiety, which is stable in the vacuole, released as the result of GFP-Atg8 degradation. (B) The amount of Atg8 recruited to the PAS is limited by the protein level. Yeast cells were starved in nitrogen-starvation medium for 1 h. Microscopy observation was performed as in Figure 3. Average peak intensities during dynamic cycles are shown. The average peak intensity in strain GA8 is normalized as 100%. Error bar, SEM; n > 100. Lower amounts of GFP-Atg8 were recruited to the PAS in strain A81-GA8 compared with strain GA8. (C) The level of autophagy correlates with the amount of GFP-Atg8 recruited to the PAS. The Pho8Δ60 assay was performed as in Figure 1. Error bar, SEM from three independent repeats.

Atg8 does not control the frequency of autophagosome production. (A) GFP-Atg8 was expressed alone (strain GA8) or with additional Atg8 (strain GA8+A8). Samples were prepared as in Figure 5A. Protein extracts were analyzed by immunoblotting with anti-GFP antibody or anti-Atg8 antiserum. (B) Frequencies of autophagosome production in strains GA8 and GA8+A8 are comparable; the atg27Δ strain showed a significantly lower frequency. Yeast cells were starved for 1 h. The microscopy observation was performed as in Figure 3. The numbers of GFP-Atg8 peaks during 15 min were recorded. Error bar, SEM; n > 100. (C) Autophagy levels are different in strains GA8 and GA8+A8. The Pho8Δ60 assay was performed as in Figure 1. Error bar, SEM from three independent repeats.