Abstract
The yeast Atg8 protein and its paralogs in mammals, mammalian Atg8s (mAtg8s), have been primarily appreciated for their participation in autophagy. However, lipidated mAtg8s, including the most frequently used autophagosomal membrane marker LC3B, are found on cellular membranes other than autophagosomes. Here we put forward a hypothesis that the lipidation of mAtg8s, termed 'Atg8ylation', is a general membrane stress and remodeling response analogous to the role that ubiquitylation plays in tagging proteins. Ubiquitin and mAtg8s are related in sequence and structure, and the lipidation of mAtg8s occurs on its C-terminal glycine, akin to the C-terminal glycine of ubiquitin. Conceptually, we propose that mAtg8s and Atg8ylation are to membranes what ubiquitin and ubiquitylation are to proteins, and that, like ubiquitylation, Atg8ylation has a multitude of downstream effector outputs, one of which is autophagy.