Transcription Factor Deformed Wings Is an Atg8a-Interacting Protein That Regulates Autophagy

Abstract

LC3 (microtubule-associated protein 1 light chain 3, called Atg8 in yeast and Drosophila) is one of the most well-studied autophagy-related proteins. LC3 controls the selectivity of autophagic degradation by interacting with LIR (LC3-interacting region) motifs also known as AIM (Atg8-interacting motifs) on selective autophagy receptors that carry cargo for degradation. Although the function of Atg8 family proteins is primarily cytoplasmic, they are also enriched in the nucleus. Despite the accumulating evidence indicating the presence of Atg8 proteins in the nucleus, the mechanisms by which they are targeted to the nucleus, their interactions with nuclear components, and their nuclear role in remain poorly understood. Here, we used yeast two-hybrid screening, and we identified transcription factor Deformed wings (Dwg) as an Atg8a-interacting protein in Drosophila. Dwg-Atg8a interaction is LIR motif-dependent. We have created Dwg Y129A/I132A LIR mutant flies and shown that they exhibit elevated autophagy, improved resistance to oxidative stress, and starvation. Our results provide novel insights into the transcriptional regulation of autophagy in Drosophila.

Keywords: LIR motif; autophagy; transcription factors.

Figure 1

(A) Yeast two-hybrid data showing interaction of the Selected Interaction Domain fragment (SID) in yellow in the region between 1 and 180 of Dwg with Atg8a. Using the online iLIR software, an LIR motif with the sequence EEYVMI was identified within the SID region at position 127–132. (B) GST pull-down assay confirming interaction between Dwg and Atg8a^WT and showing a reduced interaction between Dwg and Atg8a^LDS mutant. Experiment was performed in triplicate and data were analyzed by one-way ANOVA with a post hoc Tukey test, *** p < 0.001, ns: not significant based on the statistical analysis performed. (C) LIR motif consensus sequences of known Drosophila LIR motif-containing proteins (LIRCPs). (Left panel) Alignment of the experimentally verified functional LIR motifs identified in Drosophila LIRCPs. The LIR peptide extends six amino acids N-term and nine amino acids C-term from the Atg8a LDS HP1-binding residue (designated as position “0”). Highlighted red dotted boxes denote the critical amino acids at positions “X0” and “X3” of the LIR core that dock to the HP1 and HP2 sites, respectively, within the Atg8a LDS pocket. Acidic amino acids are represented with red, basic amino acids with blue, and with green are serine and threonine residues that may be targets of de/phosphorylation cycles. (Right panel) Sequence logo map for the conservation of each residue with respect to its position in the extended LIR motif. Sequence logo graph constructed using WebLogo (available at: https://weblogo.berkeley.edu/). (D) GST pull-down assay looking at the interaction of Dwg with Atg8a when a mutation has been introduced in the predicted LIR motif responsible for the Atg8a binding. Alanine substitutions were introduced in the predicted binding sequence at positions 129 and 132 (EEYVMI EEAVMA). Data were analyzed by one-way ANOVA with a post hoc Tukey test, n = 3, *** p < 0.001, ** p < 0.01, ns: not significant based on the statistical analysis performed.

Figure 2

(A) Superimposition of Ref(2)P (left) and Dwg (right) protein backbones docking to the HP1 (yellow) and HP2 (green) sites of Atg8a. (B) (i) Three-dimensional superimposition of Atg8a HP1 (orange) and HP2 (green) sites. (ii) Dwg protein backbone (blue) and the predicted LIR motif (red) shown docking to the LDS of Atg8a. Tyrosine seen docking in the HP1 and isoleucine docking to the HP2 pockets. (iii) Dwg protein backbone (blue) and the predicted LIR motif with two alanine mutations introduced to the sequence. The models show a lost interaction and lack of docking to the HP1 and HP2 sites. (C) Confocal micrographs of larval fat bodies under fed conditions expressing mCherry-Atg8a and stained for endogenous Dwg (green). Scale bar: 10 microns. (D) Confocal micrographs of larval fat bodies under starved conditions (4 h, 20% sucrose) expressing mCherry-Atg8a and stained for endogenous Dwg (green). Arrows show colocalization between cytoplasmic Dwg puncta and the autophagic marker mCherry-Atg8a. Scale bar: 10 microns.

Figure 3

(A) Atg8a lipidation Western blot for Atg8a-I and Atg8a-II using GABARAP in 3-week-old adult female flies. Quantification for normalized levels of Atg8a-II in Dwg LIR^Y129A/I132A mutant flies. Actin used as loading control. n = 3, * = p ≤ 0.1. One sample T test and Wilcoxon post hoc test conducted. Error bars represent SD. (B) Western blot of the same samples showing Ref(2)P levels in WT, Atg8a, and Dwg LIR^Y129A/I132A mutant flies. Data were analyzed by one-way ANOVA with a post hoc Tukey test, n = 3, **** p < 0.0001, ns: not significant based on the statistical analysis performed. (C) Dwg LIR^Y129A/I132A mutant flies show a higher resistance to oxidative stress. Survival of ~125 female flies of each genotype with 30 mM paraquat-supplemented food. Survival graph showing WT (red), Dwg LIR^Y129A/I132A (yellow), Dwg⁸/Dwg⁸(green), and Atg8a (blue). Log-rank statistical analysis was performed. WT vs. Atg8a **** p < 0.0001, WT vs. Dwg⁸/Dwg⁸ **** p < 0.0001, WT vs. Dwg LIR^Y129A/I132A **** p < 0.0001. (D) Dwg LIR^Y129A/I132A mutant flies show a higher resistance to starvation. Survival of ~125 female flies of each genotype on water-only diet. Log-rank statistical analysis was performed. WT vs. Atg8a **** p < 0.0001, WT vs. Dwg⁸/Dwg⁸ **** p < 0.0001, WT vs. Dwg LIR^Y129A/I132A **** p < 0.0001. (E) Dwg LIR^Y129A/I132A mutant flies show no change in lifespan compared to WT. Survival of ~150 female flies of each genotype was tracked over time and analyzed using a log-rank test. Female survival graph showing WT (red), and Dwg LIR^Y129A/I132A (yellow). WT vs. Dwg LIR^Y129A/I132A * p < 0.5 with a significance value of 0.013.