Autophagosome Formation: Tracing the Source

Abstract

Although it is well appreciated that autophagy begins with phagophore formation and expansion through lipids acquisition to become the autophagosome, this process remains poorly understood, with the source of autophagosome membrane controversial. Reporting recently in Nature, Hamasaki et al. (2013) suggests the ER-mitochondria contact sites are involved in phagophore assembly.

Autophagy is a degradative process during which portions of the cytoplasm are engulfed within a double-membrane compartment, the autophagosome, which ultimately fuses with the lysosome or vacuole, allowing the cargo to be degraded and recycled. Morphologically, autophagy starts with the nucleation of the phagophore, the initial sequestering organelle. The phagophore expands through the acquisition of lipids and ultimately seals, thus completing the formation process and generating an autophagosome. How the phagophore forms remains one the most fundamental unresolved questions in the understanding of the autophagy pathway. In recent years much effort has been focused on this issue, and independent studies have pointed to several different organelles as potential membrane sources. These include the plasma membrane, the Golgi apparatus, the endoplasmic reticulum (ER), and the mitochondria. In a recent study in Nature, Hamasaki et al. (2013) now shows that autophagosomes form at ER-mitochondria interfaces in mammalian cells.

Upon autophagy induction, ATG14, an autophagy-specific member of the phosphatidylinositol 3-kinase complex, is rapidly recruited to the phagophore where it plays a role in early steps of autophagosome biogenesis. Under autophagic conditions, ATG14 and phosphatidylinositol-3-phosphate are enriched in specific ER subdomains from which cup-shaped structures called omegasomes emerge (Axe et al., 2008). As these compartments are positive for components of the autophagic machinery, it was proposed that omegasomes constitute the platform for autophagosome formation.

Using immunoelectron microscopy and subcellular fractionation, Hamasaki and coworkers (2013) now show that ATG14-positive puncta assemble at the mitochondria-associated ER membrane (MAM) under starvation conditions. ZFYVE1/DFCP1, a marker of the omegasome, also shifts to the MAM upon starvation, whereas ATG5, a component of an ubiquitin-like conjugation system that is critical for autophagosome biogenesis, localizes to the ER-mitochondria junction sites during the formation process and dissociates from it after autophagosome completion. During the entire process, ATG5 shows a stable association with the ER but an oscillation in its localization at the mitochondria, suggesting that dynamic ER-mitochondria association could be required for autophagy. This idea was further tested by disrupting the ER-mitochondria contact sites, which resulted in severe impairment of ATG14 puncta formation and reduction of autophagic activity, suggesting that the relocalization of ATG14 to the MAM is required for autophagosome formation. Together, this study proposes a model in which the ER constitutes the platform for autophagosome formation, and that exchange(s) between the ER and the mitochondria are necessary for this process. The importance of the ER-mitochondria contact sites in autophagy was already pointed out in a previous study, which proposed that mitochondria are the platform for autophagosome formation (Hailey et al., 2010). The work by Hamasaki et al. (2013) thus reconciles two conflicting hypotheses: ER or mitochondria as the origin of the autophagosomal membrane.

Why are the ER-mitochondria junctions important for autophagosome biogenesis and what kind of exchange could they mediate? Mitochondrial proteins were previously noted on autophagosomes. Therefore, MAMs might mediate the incorporation of proteins required for autophagosome formation from both the ER and the mitochondria. ER-mitochondria contact sites also constitute platforms for lipid synthesis and lipid exchange between the two organelles (Rowland and Voeltz, 2012). In particular, the mitochondrial synthesis of phosphatidylethanolamine (PE), a critical lipid for autophagy, uses phosphatidylserine (PS) provided by the ER. As assessed by a fluorescent probe, incorporation of lipids from the ER into the mitochondrial membrane and subsequently into the autophagosome has been detected (Hailey et al., 2010). Hence, ER-mitochondria contacts could mediate the synthesis and/or trafficking of lipids required for autophagosome formation. It is noteworthy that the ER-mitochondria contact sites are defined as regions of close proximity, but not fusion, between the membranes of the two organelles. Previous independent studies have reported membrane continuity between the ER or the mitochondria, and the phagophore. Could the phagophore membrane fuse with the two organelles at the MAM? Otherwise, how are components delivered to the forming autophagosomes from the mitochondria and/or the ER?

The study by Hamasaki and collaborators (2013) provides compelling evidence supporting the idea that autophagosomes form at the ER-mitochondria interface. Nevertheless, this does not exclude the possibility of other sites contributing membrane for autophagosome formation. 3D tomographic studies show that 30% of the autophagosome precursors are not associated with the ER (Hayashi-Nishino et al., 2009), suggesting that there are other sites/membrane sources for autophagosome formation. Several lines of evidence have led to the proposal that the plasma membrane as well as the Golgi apparatus provide lipids to the phagophore (for a review see Mari et al., 2011). Together, these studies suggest that autophagosomes are composed of membrane from multiple sources. Multiple membrane sources for autophagosome formation could be required to maintain autophagy during prolonged periods of starvation without affecting the homeostasis of any particular individual subcellular compartment. Another possibility is that different types of autophagy induction lead to the formation of different types of autophagosomes. This latter possibility could explain a discrepancy between data from Hamasaki et al. (2013) regarding STX17, an ER- and mitochondria-associated Qa SNARE, and that from another recent independent study (Itakura et al., 2012). Whereas, Hamaski et al. (2013) show that STX17 is required for the shift of ATG14 to the MAM, and further demonstrate a direct interaction between STX17 and ATG14, STX17 was observed in the other work to translocate only to completed autophagosomes, mediating their fusion with endosomes or lysosomes, without colocalization with ATG14 (Itakura et al., 2012). Could different autophagy-inducing conditions account for these different roles for STX17?

Although questions remain to be answered before we can gain a complete understanding of the process of phagophore biogenesis and decisively resolve the membrane sources for autophagosome formation, the study by Hamasaki and collaborators (2013) provides valuable steps forward in solving the puzzle.