Selective autophagy is a protein targeting process that captures large unwanted structures into double-membrane transport vesicles called autophagosomes, which deliver the trapped targets to the lysosome for degradation (Figure 1A). Over the past two decades, the budding yeast S. cerevisiae has provided many critical mechanistic insights into this process, as well as the related process of starvation-induced non-selective autophagy, starting with the identification of Atg proteins dedicated to the execution of autophagy from pioneering genetic screens; more recently studies of Atg homologs in human cells have revealed the deep conservation of selective autophagy in eukaryotic cell evolution. Despite these discoveries, many of the most fundamental questions about selective autophagy regulation have not been resolved: What types of organelle damage can be detected by the autophagy machinery? How do Atg proteins make precise, local decisions that decide the fate of individual organelles in a cellular population of organelles (rather than average decisions about the organelle population as a whole)?
Regulation of selective autophagy
Two aspects of autophagosome construction are subject to regulation: 1) the set of targets selected for autophagosomal sequestration and 2) the rate and subcellular location of autophagosome initiation. A large body of work has shown that a class of proteins called autophagy receptors dictates target selection by tethering targets to the membrane of a forming autophagosome (Figure 1B). A second body of work has shown that the rate and location of autophagosome initiation is controlled by the Atg1/ULK1 kinase, which phosphorylates several Atg proteins to drive the formation of a new autophagosome (Figure 1C). Up-regulation of Atg1/ULK1 kinase activity by starvation enables increased production of non-selective autophagosomes but a connection between selective autophagy targets in kinase activity had not been made.
We have recently found that autophagy receptor proteins also control the activation of Atg1 during selective autophagy in nutrient-rich conditions (Figure 1D). The mechanism we described provides an explanation for how autophagy signals that are generated locally by a damaged organelle or a protein aggregate can drive destruction of individual targets. More broadly, our findings highlight that autophagy receptor proteins are at the core of signaling hubs that make selective autophagy decisions. This work transforms the big but diffuse questions about selective autophagy regulation into a manageable series of increasingly more precise mechanistic questions about autophagy receptor function.
Many autophagy receptors reside constitutively at the surface of their targets and are regulated entirely by post-translational modifications – predominantly phosphorylation. Several kinase-receptor pairs have been identified, however, little is known about how autophagy-inducing stimuli bring about receptor phosphorylation by general cytosolic kinases. Fundamentally, it remains unclear if the signaling pathways in control of surface-bound receptor phosphorylation can respond to local cues for receptor activation. Our goal is to dissect novel mechanisms by which damaged yeast organelles signal their destruction.
Hunting for novel selective autophagy factors in mammalian cells
Over the past decade, it has become clear that selective autophagy plays an important role in protecting the nervous system, which comprises post-mitotic cells that cannot dilute damaged cytosolic structures through growth and cell division. With increased lifespan, Americans have also started suffering more from neurodegenerative diseases and the development of novel therapies for treating these diseases is becoming a national priority. It is known that the efficiency of selective autophagy as a protein targeting mechanism declines with human age; but promisingly, re-activation of autophagy in a mouse aging model can ameliorate neurodegeneration and reverse senescent phenotypes.
Given that most described autophagy mechanisms tend to be conserved, the power of yeast as a model organism for understanding mammalian selective autophagy will continue to endure. Nonetheless, we started directly studying selective autophagy in mammalian cells a couple of years ago because there are many gaps in our understanding of selective autophagy mechanisms that are present in mammals but absent from yeast. How do we fill them in? Until very recently the idea of doing selective autophagy screens in mammalian cells to find missing components without clear yeast homologs or equivalents was considered fantasy. That has all changed recently with the developments in CRISPR/Cas9 technology. However, this technology is enabling but not sufficient by itself to revolutionize the study of mammalian selective autophagy because it also necessitates quantitative phenotypic readouts of mutant phenotypes. This has prompted us to devise new reporters of receptor function that will facilitate subsequent genetic screening.