Beyond Atg8 binding: The role of AIM/LIR motifs in autophagy

ABSTRACT

Selective macroautophagy/autophagy mediates the selective delivery of cytoplasmic cargo material via autophagosomes into the lytic compartment for degradation. This selectivity is mediated by cargo receptor molecules that link the cargo to the phagophore (the precursor of the autophagosome) membrane via their simultaneous interaction with the cargo and Atg8 proteins on the membrane. Atg8 proteins are attached to membrane in a conjugation reaction and the cargo receptors bind them via short peptide motifs called Atg8-interacting motifs/LC3-interacting regions (AIMs/LIRs). We have recently shown for the yeast Atg19 cargo receptor that the AIM/LIR motifs also serve to recruit the Atg12–Atg5-Atg16 complex, which stimulates Atg8 conjugation, to the cargo. We could further show in a reconstituted system that the recruitment of the Atg12–Atg5-Atg16 complex is sufficient for cargo-directed Atg8 conjugation. Our results suggest that AIM/LIR motifs could have more general roles in autophagy.

The selectivity of macroautophagic processes involves the specific interaction of cargo binding receptor molecules, which link the cargo to the nascent autophagosomal membrane (i.e., the phagophore) via their interaction with Atg8 proteins. Atg8 proteins are ubiquitin-like proteins that are covalently attached to the membrane lipid phosphatidylethanolamine (PE). This conjugation reaction requires the action of the E1- and E2- like enzymes Atg7 and Atg3, respectively, and is enhanced by the Atg12–Atg5-Atg16 complex that acts in an E3-like manner. The cargo receptors interact with Atg8 proteins via short peptide motifs called AIMs in yeast or LIRs in more complex eukaryotes (also referred to as LC3 recognition sequences/LRSs).

In our recent paper we have made steps toward a better understanding of how Atg8 conjugation can be coupled to the presence of the cargo and, furthermore, toward the role of AIM/LIR motifs in selective autophagy. In particular, we found that the S. cerevisiae cargo receptor Atg19 directly binds to the Atg5 subunit of the Atg12–Atg5-Atg16 complex (Fig. 1). This interaction is mediated by the AIM/LIR motifs of the Atg19 cargo receptor, which bind to 2 pockets in Atg5. Due to this interaction Atg19 is able to recruit the Atg12–Atg5-Atg16 complex to the cargo, stimulating local Atg8 conjugation to PE (Fig. 1). To our knowledge this is the first description of the requirements for LIR motifs for protein interactions with autophagic components other than Atg8 proteins.

Figure 1.

The role of AIM/LIR motifs in Atg8 conjugation at the cargo. The prApe1 cargo (a spherical representation with only a few prApe1 dodecamers shown for simplicity) is bound by its cargo receptor Atg19. The C-terminal AIM/LIR motifs of Atg19 interact with the E3-like Atg12–Atg5-Atg16 complex via the Atg5 subunit and thereby recruit it close to the cargo. This in turn stimulates local Atg8–PE conjugation mediated by the E1-like (Atg7) and E2-like (Atg3) enzymes. The membrane-bound Atg8 may eventually outcompete the Atg12–Atg5-Atg16 complex. The high avidity interaction of the AIM/LIR motifs in Atg19 with Atg8 could subsequently result in close membrane-cargo apposition.

In the well-studied selective autophagy-like cytoplasm-to-vacuole targeting (Cvt) pathway in S. cerevisiae the cargo receptor Atg19 mediates the selective transport of precursor aminopeptidase I (prApe1) oligomers into the vacuole. Atg19 uses up to 3 AIM/LIR motifs (the canonical ₄₁₂WEEL₄₁₅ at its extreme C terminus and 2 cryptic ones more upstream), for its binding to Atg8 and thereby mediates selective delivery of the cargo. While studying this pathway, we found that the cargo receptor Atg19 interacts directly with the Atg5 subunit of the Atg12–Atg5-Atg16 complex and that this interaction requires the AIM/LIR motifs in Atg19. Molecular dynamics simulations suggested that the interactions of the AIM/LIR motifs might occur with 2 distinct sites of the Atg5 protein. The architecture of both these binding sites consists of hydrophobic pockets surrounded by positively charged residues on the surface of the protein, resembling the structural and chemical characteristics of AIM/LIR-binding pockets in Atg8 proteins. Mutagenesis resulting in reversion of the positive charges abolishes the Atg19-Atg5 interaction in vitro and strongly impairs the interaction in vivo as determined by co-immunoprecipitation experiments. Although Atg8 and Atg5 share similar secondary and tertiary structural features (i.e., ubiquitin-like folds) superposition of the Atg5 and Atg8 structures shows that the AIM/LIR binding sites of the 2 molecules occupy different positions in relation to the ubiquitin-like domains.

To further analyze the interaction between Atg19 and Atg12–Atg5-Atg16 in the context of Atg8 conjugation to PE, we developed an in vitro reconstitution system using cargo mimetic beads coated with the purified prApe1 propeptide and the Atg19 cargo receptor. The E3-like Atg12–Atg5-Atg16 complex was recruited to these cargo mimetic beads and upon addition of the Atg8 conjugation machinery, composed of Atg3, Atg7 and Atg8 itself, and in the presence of PE-containing small unilamellar vesicles and ATP, we could reconstitute Atg8 conjugation at the beads. This result shows that the recruitment of the Atg12–Atg5-Atg16 complex to the prApe1 cargo via Atg19 is sufficient to drive cargo-directed Atg8 conjugation. The Atg12–Atg5-Atg16-Atg19 interaction is mutually exclusive with the binding of Atg8 to the same AIM/LIR motifs in Atg19 suggesting a hierarchy of binding events where the end product of the catalytic conjugation reaction has the strongest interaction with the receptor (Fig. 1). This model is in line with the finding that the AIM/LIR-mediated high avidity interaction of Atg19 with multiple Atg8 molecules concentrated on the nascent Cvt vesicle membrane leads to tight apposition of the membrane and the cargo. Hence the AIM/LIR-mediated interaction of Atg19 with Atg5 and Atg8 might contribute a mechanism to ensure directionality to the process of phagophore membrane elongation.

Moreover, we hypothesize a certain degree of conservation of this mechanism since we also observed binding of the mammalian SQSTM1/p62, OPTN and CALCOCO2/NDP52 cargo receptors to ATG5. The molecular mechanisms of this interaction still need to be elucidated and it is currently unclear if the binding between these molecules is direct or mediated by another protein(s).

Overall, it appears that AIM/LIR motifs could play more general roles during autophagy by binding to factors other than Atg8 proteins. AIM/LIR motifs have been described in many autophagy-related proteins including the components of the Atg8 conjugation machinery (i.e., mammalian ATG4 and yeast Atg3) and Atg1, and the AIM/LIR motif might therefore be a more general mechanism to concentrate autophagy proteins at the site of autophagosome formation. An additional level of regulation as to which factor is bound by the AIM/LIR motifs might occur at the level of post-translational modifications. One of which may be phosphorylation (for example in OPTN) as a large number of the known AIM/LIR motifs harbor a serine or threonine either between the hydrophobic residues or in front of the first hydrophobic residue. In summary, although we have begun to get insights into the intricacy of AIM/LIR-mediated interactions in autophagy, we still miss a more detailed picture of how the timing and location of these interactions are orchestrated. Future work will bring to light more AIM/LIR holders and AIM/LIR recipients, and put them into play to build a fine regulatory network in the orchestration of autophagy.