Drosophila p53 integrates the antagonism between autophagy and apoptosis in response to stress

Abstract

The tumor suppressor TP53/p53 is a known regulator of apoptosis and macroautophagy/autophagy. However, the molecular mechanism by which TP53 regulates 2 apparently incompatible processes remains unknown. We found that Drosophila lacking p53 displayed impaired autophagic flux, higher caspase activation and mortality in response to oxidative stress compared with wild-type flies. Moreover, autophagy and apoptosis were differentially regulated by the p53 (p53B) and ΔNp53 (p53A) isoforms: while the former induced autophagy in differentiated neurons, which protected against cell death, the latter inhibited autophagy by activating the caspases Dronc, Drice, and Dcp-1. Our results demonstrate that the differential use of p53 isoforms combined with the antagonism between apoptosis and autophagy ensures the generation of an appropriate p53 biological response to stress.

Keywords: Apoptosis; Parkinson disease; caspase; macroautophagy; neurodegenerative disease; oxidative stress; p53; photoreceptor.

Figure 1.

p53- and autophagy-mutant flies show increased sensitivity to paraquat. (a and b) Survival curves of wild-type and mutant Drosophila fed with normal or 20 mM paraquat (PQ)-containing media (A, n = 4 and B, n = 5). 1) Control, w1118 flies, 2) p53^null (p53^5A1−4 mutants), 3) p53BAC, p53^null flies carrying a wild-type p53 genomic rescue BAC clone on the second chromosome and 4) atg8^−/-, (atg8^KG07569 mutants). (c) Densitometric quantification of endogenous ref(2)P levels in whole flies fed with normal or PQ-containing media. Data are the mean ± SEM of 5 independent blots and are presented as the ratio of ref(2)P:tubulin in experimental animals relative to the ratio in control non-treated flies. *P < 0.05, **P < 0.01 by one-way ANOVA followed by the Bonferroni post hoc test. (d) Quantification of Atg8-II:Atg8-I from western blots of control and p53^null flies treated or non-treated with PQ. Data are the mean ± SEM of n = 3 blots and are presented as expression levels relative to that in non-treated control flies.

Figure 2.

p53- and autophagy-defective flies show increased caspase activation after paraquat treatment. (a and b) Representative western blot and densitometric quantification of endogenous cleaved Dcp-1 in whole flies. Data are the mean ± SEM of 3 independent blots and are expressed as the ratio of cleaved Dcp-1:tubulin. ***P < 0.001 by one-way ANOVA followed by the Bonferroni post hoc test. (c) Dorsomedial protocerebrum area of the fly brain selected for quantification of immunoreactivity to an anti-cleaved human CASP3 antibody (yellow box). (d–g) Representative images of the above-mentioned cleaved human CASP3 immunostaining in the dorsomedial protocerebrum of untreated control (d) and p53^−/- (e) flies or PQ-treated control (f) and p53^−/- (g) flies. Scale bars: 25 µm. White arrows show cells positively stained for cleaved human CASP3. (h) Quantification of staining with immunoreactivity against cleaved human CASP3 in the dorsomedial protocerebrum from untreated (n = 8) and PQ-treated (n = 6) control flies, or untreated (n = 5) and PQ-treated (n = 5) p53^−/- flies. Data are the mean ± SEM of the number of dots of immunoreactivity against cleaved human Caspase 3 counted in the dorsomedial protocerebrum. **P < 0.01, ***P < 0.001 by one-way ANOVA followed by the Bonferroni post hoc test.

Figure 3.

Differential ref(2)P processing in photoreceptors expressing p53 (p53B) or ∆Np53 (p53B). (a–f) Representative fluorescence microscopy images of the retinas of adult flies expressing autophagic flux reporter GFP-ref(2)P in photoreceptors (PRs). The nuclei were stained with DAPI (blue). Flies also overexpressed mCD8-RFP (control protein), ∆Np53 (p53A), p53 (p53B), gug/atro75QN, ∆Np53 (p53A) and p35, or p53 (p53B) and p35 in PRs under the control of ninaE/rh1-GAL4. Actin was visualized by staining with phalloidin (red). Scale bars: 20 µm. (g) Quantification of GFP-ref(2)P dots per retina in the strains represented in (a–f) using the Find Maxima function of ImageJ software. Data are the mean ± SEM of n = 8 to 15 retinas. *P < 0.05, **P < 0.01 by the Student t test.

Figure 4.

Differential processing of mCherry-GFP-Atg8 in p53 (p53B)- and ∆Np53 (p53A)-expressing flies. (a–f). Representative fluorescence microscopy images of the retina of adult flies overexpressing the autophagy reporter mCherry-GFP-Atg8 plus LacZ (control) ± p35, ∆Np53 (p53A) ± p35, and p53 (p53B) ± p35 driven by ninaE/rh1-GAL4. mCherry in red and GFP is green. Staining from n = 2 to 10 retinas. Arrowheads show high levels of GFP-positive dots. Scale bars: 10 μm. (g) Quantification of colocalization of cherry and GFP staining by the Pearson coefficient from A to F, n = 15–26 retinas *P < 0.05, **P < 0.01 by one way Anova followed by the Tukey multiple comparison test.

Figure 5.

Inhibition of Dronc, Drice, or Dcp-1 restores ref(2)P processing in photoreceptors expressing ∆Np53 (p53A). (a–f) Representative fluorescence microscopy images of the retinas of adult flies expressing the autophagic flux reporter GFP-ref(2)P in photoreceptors (PRs). Flies also overexpressed mCD8-RFP (control protein), ∆Np53 (p53A), p53 (p53B), gug/atro75QN, ∆Np53 (p53A) and dronc^I29, or p53 (p53B) and dronc^I29 in PRs under the control of ninaE/rh1-GAL4. Dronc^I29 flies carry one dronc^I29 mutant allele. Actin and cell nuclei were visualized by staining with phalloidin (red) and DAPI (blue), respectively. Scale bars: 20 µm. (g) Quantification of GFP-ref(2)P dots per retina area in the strains represented in (a–f). Data are the mean ± SEM of n = 8 to 17 retinas. *P < 00.5 by the Student t test. (h–o) Representative images of the retina of adult flies expressing the autophagic flux reporter GFP-ref(2)P in PRs. Flies also overexpressed mCD8-RFP (control), ∆Np53 (p53A), ∆Np53 (p53A) with dcp-1RNAi, ∆Np53 (p53A) with drice-1RNAi, gug/atro75QN, p53 (p53B), p53 (p53B) with dcp-1RNAi, or p53 (p53B) with drice-1RNAi under the control of ninaE/rh1-GAL4. Actin and cell nuclei were visualized by staining with phalloidin (red) or DAPI (blue), respectively. Scale bars: 20 µm. (p) Quantification of GFP-ref(2)P dots per retina area in the strains represented in (h–o). Data are the mean ± SEM of n = 8 to 15 retinas. *P < 0.05 by the Student t test.

Figure 6.

atg1 mutant photoreceptors show increased sensitivity to p53 (p53B)-induced, but not ∆Np53 (p53A)-induced death. Ommatidia from flies overexpressing LacZ (control), ∆Np53 (p53A), or p53 (p53B) in PRs under the control of ninaE/rh1-GAL4. (a) Representative mosaic ommatidia with wild-type (ey-flp, ninaE/rh1-GAL4;UAS-GFP; FRT80 atg1Δ3D/FRT80 ninaE/Rh1-tomato) and atg1^−/- mutant (ey-flp, ninaE/rh1-GAL4;UAS-GFP; FRT80 atg1Δ3D/FRT80 atg1Δ3D) PRs, as generated for (b–d). Numbers 1 to 6 indicate the outer PRs (R1–6). Wild-type PRs carry 2 copies of Tomato in addition to GFP and thus appear yellow Mutant clones, in green, were generated by mitotic recombination and thus carry GFP but not the Tomato reporter. (b–d) Representative retinas showing atg1∆3D mutant clone size). Scale bars: 20 µm. (e) Quantification of PR loss in wild-type and mutant clones from n = 12 to 14 retinas. Data are shown as global PR loss per fly strain. ***P < 0.001 by the Student t test.

Figure 7.

Life and death: levels of p53 isoforms matter. (Left) When p53 isoforms are overexpressed in photoreceptor neurons at high levels, p53 (p53B) induces autophagy and caspase-independent cell death, while ∆Np53 (p53A) induces caspase activation and inhibits autophagic flux. (Right) Physiological levels of p53 isoforms confer redundant resistance to reactive oxygen species (ROS) and a functional autophagic flux. Pink schematic labeled photoreceptors on the left represents one ommatidium containing 8 photoreceptor neurons and accessory cells.