Exploring non-canonical functions of human ATG8 proteins

• Applicant: Professor Christian Behrends, Ph.D.
• Subject Area: Cell Biology ; Biochemistry
• Term: since 2016
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 318126409

The ATG8 family proteins are critical players in autophagy, a conserved degradation pathway which delivers defective und unneeded cytosolic material to lysosomes. During autophagy, ATG8s are conjugated to double-membrane autophagosomes to facilitate cargo recruitment, autophagosome biogenesis, transport and fusion with lysosomes. In addition, recent reports demonstrate that ATG8s are also delivered to single-membrane organelles with highly divergent degradative or secretory outcomes. In these processes, ATG8 proteins exert non-canonical functions whose regulatory mechanisms remain to be fully elucidated. However, the reliance on overexpression and the lack of specific antibodies has hampered the analysis of these emerging functions. In the past funding period, we employed genome-editing to generate a panel of cells in which individual LC3 and GABARAP genes of the ATG8 family were seamlessly tagged at their natural chromosomal locations with an affinity epitope. These cells have been distributed across the autophagy community and proven to be a valid resource for studying GABARAPs and LC3s similar than ATG8 family knockouts and peptide-based sensors which have been independently established by other groups. Using these cells, we recently mapped endogenous GABARAPL2 protein complexes. This approach identified the ER-associated and lipid droplet (LD) biogenesis factor ACSL3 as a stabilizing GABARAPL2-binding partner. GABARAPL2 bound ACSL3 in a manner dependent on its LC3-interacting regions, whose binding site in GABARAPL2 was required to recruit the latter to the ER. Importantly, the interaction between GABARAPL2 and ACSL3 contributed to anchor the protein UBA5 at the ER. As an E1-like enzyme, UBA5 is part of the enzymatic cascade whose activity leads to the covalent attachment of the ubiquitin-like protein UFM1 to target proteins (i.e., ufmylation). Consistent with the notion that ACSL3 helps to spatially assembly the ufmylation machinery, ACSL3 depletion or LD formation affected the abundance of UFM1 conjugation components and inhibited ER-phagy which is a known regulatory target of these factors. Hence, we identified the GABARAPL2-ACSL3 interaction as novel regulator of ufmylation. Consequently, in the present proposal we propose to further investigate this new non-canonical role of GABARAPL2 by studying its binding to ACSL3, by determining the ER anchoring mechanism of UBA5 and by identifying targets of the ufmylation reaction in response to lipid stress. Our discovery that induction of LD formation is a prominent inducer of ufmylation provides a powerful mean to overcome past difficulties of exploring the ufmylome due to the lack of defined conditions that trigger UFM1 conjugation. Importantly, UFM1 target identification is essential for understanding the mechanism of this unusual but essential ubiquitin-like conjugation system in general and for dissecting the biological significance of coupling ufmylation and LD biogenesis in particular.

• DFG Programme: Research Grants