Reactive oxygen species are essential for autophagy and specifically regulate the activity of Atg4

Abstract

Autophagy is a major catabolic pathway by which eukaryotic cells degrade and recycle macromolecules and organelles. This pathway is activated under environmental stress conditions, during development and in various pathological situations. In this study, we describe the role of reactive oxygen species (ROS) as signaling molecules in starvation-induced autophagy. We show that starvation stimulates formation of ROS, specifically H(2)O(2). These oxidative conditions are essential for autophagy, as treatment with antioxidative agents abolished the formation of autophagosomes and the consequent degradation of proteins. Furthermore, we identify the cysteine protease HsAtg4 as a direct target for oxidation by H(2)O(2), and specify a cysteine residue located near the HsAtg4 catalytic site as a critical for this regulation. Expression of this regulatory mutant prevented the formation of autophagosomes in cells, thus providing a molecular mechanism for redox regulation of the autophagic process.

Figure 1

Cells accumulate ROS under starvation conditions. (A) CHO cells were grown in a control medium (see Materials and methods) or starved for 3 or 13 h, after which they were incubated in 50 μM DHE and visualized as detailed in Materials and methods. (B) CHO cells were grown as in (A), treated with 5 nM MitoTracker Red for 30 min at 37°C, washed, treated with 30 μM DCFDA and visualized as detailed in Materials and methods. (C) HeLa cells grown in 96-well plates were deprived of serum (αMEM), completely starved (EBSS) or maintained in a control medium (α-MEM, 10% FCS) for 2 h, after which they were treated with DCFDA as in (B) and subsequently analyzed in a fluorimeter, as explained in Materials and methods. (D) Data collected from the fluorometric measurements were analyzed as detailed in Materials and methods.

Figure 2

ROS formation occurs partially downstream of class III PI3K activation. (A) Left panel: HeLa cells were starved for 2 h in the presence or absence of either 100 nM wortmannin or 10 mM 3MA, after which they were treated with DCFDA and analyzed by a fluorimeter as described above. Right panel: the rate of degradation of long-lived proteins was measured in cells incubated in either α-MEM medium or EBSS medium in the absence or presence of 100 nM wortmannin or 10 mM 3MA. Values are represented as the means±s.d. of three separate determinations. (B) WT MEFs or Atg5 (−/−) MEFs from two separate clones were starved for 2 h before treatment with DCFDA and visualization or fluorometric analysis.

Figure 3

ROS accumulation is essential for autophagy. (A) Upper panel: CHO cells stably transfected with GFP-GATE-16 were preincubated in the presence or absence of 10 mM NAC or 1000 μ/ml catalase for 10 min before starvation for 2 h in the presence or absence of these drugs, or grown in a control medium containing the drugs for 2 h. The cells were then fixed, permeabilized and stained with anti-GFP monoclonal antibodies. Representative images are shown. Lower panel: HEK 293 cells were transfected with GFP-GATE-16. At 24 h post-transfection, the cells were treated with NAC or catalase and starved as explained above, lysed in Ripa buffer and 100 μg of each lysate was separated on 10% SDS–PAGE and subsequently analyzed with anti-GFP antibodies to detect the transfected GATE-16 and anti-tubulin antibodies as control. The data were quantified using NIH image program and are depicted as the percentage of lipidated protein from the total GATE-16. (^*) indicates non-lipidated and (^**) indicates lipidated GFP-GATE-16. (B) CHO cells treated with NAC or catalase and starved as detailed in (A) were incubated with DCFDA and visualized by a confocal microscope or analyzed by a fluorimeter as in Figure 1. (C) The rate of degradation of long-lived proteins was measured in CHO cells incubated in either α-MEM medium or EBSS medium, or following pretreatment with 10 mM NAC for 10 min or with 1000 μ/ml catalase overnight. A representative experiment is shown.

Figure 4

The activity of endogenous Atg4 is inhibited under nutrient starvation. (A) CHO cells (for mammalian Atg8s) or cells of S. cerevisiae (for ScAtg8) were incubated in starvation medium (EBSS or SD-N, respectively) for different time periods as indicated before isolation of RNA by the Tri-reagent as explained in Materials and methods. For each gene, data obtained from non-starved cells were set to an arbitrary value of 1, and results from the corresponding starved cells were normalized accordingly. Results represent the means±s.d. of three separate experiments. (B) HEK 293 cells were grown in control medium or starved for 2 h in EBSS in the absence or presence of 100 nM bafilomycin A1. Lysates (100 μg) obtained in Ripa buffer were run on 12% SDS–PAGE and analyzed by Western blot, using anti-GATE-16 or anti-tubulin antibodies. (C) CHO cells were grown in a control medium in the presence or absence of 1 mM H₂O₂ for 1 h, or starved for 3 or 13 h. Lysates (10 μg) obtained in Ripa buffer were incubated with recombinant His₆-GATE-16-HA (0.3 μg) in 50 KT reaction buffer (25 mM Tris, pH 7.4, 50 mM KCl) at 30°C for 45 min, in the presence or absence of 1 mM DTT. The reaction was stopped by addition of sample buffer and boiling, after which the samples were resolved on 15% SDS–PAGE and subsequently analyzed by Western blot, using anti-His monoclonal antibodies. The experiment was repeated six times; a representative blot is shown. (D) Left panel: HeLa cells transfected with Myc-GATE-16-HA or with an empty vector as control were labeled with [³⁵S]methionine for 10 min and lysed immediately or chased for 1 h before lysis in Ripa buffer. Lysates were immunoprecipitated using anti-Myc antibodies and the immunoprecipitates were resolved on 15% SDS–PAGE. Middle panel: HeLa cells transfected with Myc-GATE-16-HA were kept in control medium or starved for 30 min or 13 h in EBSS before labeling with [³⁵S]methionine for 10 min, immediate lysis, analysis by SDS–PAGE and quantification (right panel) using NIH image program. Values are presented as the average percentage of unprimed form out of the total GATE-16. (^*) in all sections of this figure indicates non-cleaved His₆-GATE-16-HA or Myc-GATE-16-HA and (^**) indicates cleaved His₆-GATE-16 or Myc-GATE-16.

Figure 5

H₂O₂ directly inhibits the activity of HsAtgA. (A) Cleavage activity was tested by incubation of recombinant His₆-HsAtg4A (0.1 μg) and His₆-GATE-16-HA (0.3 μg) in 50 KT reaction buffer at 30°C for 45 min in the presence of indicated concentrations of DTT followed by Western blot analysis, using anti-His monoclonal antibodies. The resulting bands from three separate experiments were quantified with a densitometer using the Bio-Rad Multi-Analyst program and are presented as the average percentage of cleaved form out of the total GATE-16. (^*) indicates non-cleaved His₆-GATE-16-HA and (^**) indicates cleaved His₆-GATE-16. (B) His₆-HsAtg4A was incubated in the presence of 200 μM DTT at 4°C for 10 min. Reduced His₆-HsAtg4A (0.1 μg) was then incubated in 50 KT (to obtain 15 μM DTT) with the indicated concentrations of H₂O₂ at 25°C for 5 min, after which recombinant His₆-GATE-16-HA (0.3 μg) was added and incubation proceeded at 30°C for 45 min. Reaction mixtures from three separate experiments were analyzed and are presented as explained in (A). (C) His₆-HsAtg4A (0.1 μg) was incubated with recombinant His₆-GATE-16-HA (0.3 μg) after the following procedures: no treatment (lane 1); pretreatment with 200 μM DTT at 25°C for 5 min (lane 2); treatment with 200 μM DTT followed by treatment with 1 mM H₂O₂ at 25°C for 5 min (lane 3); and treatment with 200 μM DTT followed by treatment with 1 mM H₂O₂ for 5 min and then 2 mM DTT (lane 4). Reaction mixtures were analyzed by Western blot using anti-His monoclonal antibodies.

Figure 6

Tetrapod homologues of Atg4 share several conserved cysteine residues. (A) Amino-acid alignment of the region surrounding the active cysteine residue among members of the Atg4 family. Sequences of the following accession numbers are aligned: BAB83890, BAB83889, NP_777364, NP_777363, NP_001001171, CAG32326, AAH73017, AAH82660, AAH95617, XP_393739, NP_608563, XP_661074, NP_014176 and NP_191554. The alignment was preformed using the ClustalX multiple alignment program and is depicted by the ESPript 2.0 program, representing identity by a black frame and homology by a gray background. Numbering is according to the HsAtg4B sequence. α-Helices and a β-sheet found in this region according to the recently solved structure of HsAtg4B (Sugawara et al, 2005; Kumanomidou et al, 2006) are marked as black ellipses and a gray arrow, respectively. The catalytic cysteine residue is marked by a wide black arrow; other conserved cysteines are marked by thin black arrows. (B) A phylogenetic tree was created based on the alignment presented in (A) using the NJplot program (Perriere and Gouy, 1996). Bootstrap values indicating the reliability of the tree branches (with 1000 being the maximal value) are shown.

Figure 7

Mutation in Cys81 reduces the redox sensitivity of HsAtg4A in vitro. (A) Recombinant His₆-HsAtg4A^WT, His₆-HsAtg4A^C77A or His₆-HsAtg4A^C81S (0.1 μg) was incubated with His₆-GATE-16-HA (0.3 μg) in 50 KT reaction buffer at 30°C for 45 min in the presence or absence of 1 mM DTT. Reaction mixtures were analyzed by Western blot using anti-His monoclonal antibodies. (^*) indicates non-cleaved His₆-GATE-16-HA and (^**) indicates cleaved His₆-GATE-16. (B) His₆-HsAtg4A^C81S (0.1 μg) was incubated with recombinant His₆-GATE-16-HA (0.3 μg) following the same procedure as detailed in Figure 5C. Reaction mixtures were analyzed and are presented as explained in (A).

Figure 8

Cys 81 is required for redox regulation of HsAtg4A in cells. (A) CHO cells stably expressing GFP-GATE-16 were transiently transfected with HsAtg4A^WT-Myc-His₆ or HsAtg4A^C81S-Myc-His₆. At 24 h post-transfection, cells were starved for 2.5 h after which they were fixed, permeabilized and incubated with anti-Myc monoclonal antibodies. Upper panel: typical images of cells transfected with the above constructs, as visualized using a confocal microscope; representative autophagosomes are indicated by arrows. Lower panel: quantification of the average number of autophagosomes per cell in the different transfectants. Images of the fixed cells were visualized using a Nikon eclipse TE300 fluorescent microscope and used for quantification of the number of autophagosomes per cell. The results presented are the means±s.d. of a total of 100 cells from three separate experiments. (^*) indicates significance at P<0.001. (B) Left panel: HEK 293 cells were cotransfected with GFP-GATE-16 and each of the Atg4A constructs mentioned in (A). At 40 h post-transfection, the cells were starved for 2.5 h, lysed in Ripa buffer and 100 μg of each lysate was loaded to 10% SDS–PAGE and subsequently analyzed with anti-GFP antibodies to detect the transfected GATE-16, anti-Myc antibodies to detect the transfected HsAtg4A and anti-tubulin antibodies as control. (^*) indicates non-lipidated GFP-GATE-16 and (^**) indicates lipidated GFP-GATE-16. Right panel: results from three separate experiments, as detailed in the left panel, were analyzed using NIH image program and quantified as follows: the amount of lipidated GFP-GATE-16 out of the total of GFP-GATE-16 was calculated for each mutant in each experiment. The value obtained for HsAtg4A^WT in each experiment was set to 100% and the relative lipidation in cells transfected with the mutant was calculated accordingly.

Figure 9

HsAtg4B is redox regulated during starvation in a similar mechanism to HsAtg4A. (A) HeLa cells were transiently transfected with GFP-HsAtg4B^WT or GFP-HsAtg4B^C78S. At 24 h post-transfection, cells were starved for 2.5 h in the presence of 4 mM H₂O₂ after which they were fixed, permeabilized and incubated with anti-LC3 polyclonal antibodies. Cells were visualized (upper panel) and quantified (lower panel) as explained in Figure 8. Representative autophagosomes are indicated in the upper panel by arrows. The results presented in the lower panel are the means±s.d. of a total of 100 cells from three separate experiments. (^*) indicates significance at P<0.001. (B) Left panel: HeLa cells transfected and treated as in (A) were lysed in Ripa buffer and 40 μg of each lysate was loaded to 12% SDS–PAGE and subsequently analyzed with anti-LC3 antibodies and anti-GFP antibodies to detect the transfected HsAtg4B and anti-tubulin antibodies as control. (^*) indicates non-lipidated LC3 and (^**) indicates lipidated LC3. Right panel: results from three separate experiments, as detailed in the left panel, were analyzed as explained in Figure 8B.

Figure 10

A proposed model for the redox regulation of autophagy.