Human ATG4 autophagy proteases counteract attachment of ubiquitin-like LC3/GABARAP proteins to other cellular proteins.

Microtubule-associated protein 1 light chain 3 α (LC3)/GABA type A receptor-associated protein (GABARAP) comprises a family of ubiquitin-like proteins involved in (macro)autophagy, an important intracellular degradation pathway that delivers cytoplasmic material to lysosomes via double-membrane vesicles called autophagosomes. The only currently known cellular molecules covalently modified by LC3/GABARAP are membrane phospholipids such as phosphatidylethanolamine in the autophagosome membrane. Autophagy-related 4 cysteine peptidase (ATG4) proteases process inactive pro-LC3/GABARAP before lipidation, and the same proteases can also deconjugate LC3/GABARAP from lipids. To determine whether LC3/GABARAP has other molecular targets, here we generated a pre-processed LC3B mutant (Q116P) that is resistant to ATG4-mediated deconjugation. Upon expression in human cells and when assessed by immunoblotting under reducing and denaturing conditions, deconjugation-resistant LC3B accumulated in multiple forms and at much higher molecular weights than free LC3B. We observed a similar accumulation when pre-processed versions of all mammalian LC3/GABARAP isoforms were expressed in ATG4-deficient cell lines, suggesting that LC3/GABARAP can attach also to other larger molecules. We identified ATG3, the E2-like enzyme involved in LC3/GABARAP lipidation, as one target of conjugation with multiple copies of LC3/GABARAP. We show that LC3B-ATG3 conjugates are distinct from the LC3B-ATG3 thioester intermediate formed before lipidation, and we biochemically demonstrate that ATG4B can cleave LC3B-ATG3 conjugates. Finally, we determined ATG3 residue Lys-243 as an LC3B modification site. Overall, we provide the first cellular evidence that mammalian LC3/GABARAP post-translationally modifies proteins akin to ubiquitination ("LC3ylation"), with ATG4 proteases acting like deubiquitinating enzymes to counteract this modification ("deLC3ylation").

Oncogenic c-Myc induces replication stress by increasing cohesins chromatin occupancy in a CTCF-dependent manner.

Oncogene-induced replication stress is a crucial driver of genomic instability and one of the key events contributing to the onset and evolution of cancer. Despite its critical role in cancer, the mechanisms that generate oncogene-induced replication stress remain not fully understood. Here, we report that an oncogenic c-Myc-dependent increase in cohesins on DNA contributes to the induction of replication stress. Accumulation of cohesins on chromatin is not sufficient to cause replication stress, but also requires cohesins to accumulate at specific sites in a CTCF-dependent manner. We propose that the increased accumulation of cohesins at CTCF site interferes with the progression of replication forks, contributing to oncogene-induced replication stress. This is different from, and independent of, previously suggested mechanisms of oncogene-induced replication stress. This, together with the reported protective role of cohesins in preventing replication stress-induced DNA damage, supports a double-edge involvement of cohesins in causing and tolerating oncogene-induced replication stress.