Defining the membrane precursor supporting the nucleation of the phagophore

Abstract

How does the phagophore form? Which membrane acts as a platform for its biogenesis? Over the years, extensive use of microscopy techniques have led to the controversial identification of multiple potential membranes as precursors for phagophore nucleation and/or for the supply of lipids to the expanding compartment. Nevertheless, none of these studies has established a direct functional link between membrane sources and autophagosome biogenesis. Addressing this point, in a recent study highlighted by a punctum in this issue, Ge and coworkers developed an in vitro approach to determine the identity of the membranes responsible for the lipidation of LC3, thus identifying the ER-Golgi intermediate compartment (ERGIC) as a potential key determinant of phagophore biogenesis.

Upon induction, macroautophagy starts with the nucleation of the phagophore membrane that will then expand through acquisition of lipids, and ultimately seal to generate an autophagosome. The mechanism of phagophore nucleation is not known, but this structure may be generated by the sequential association of at least a subset of ATG proteins. An early step of phagophore nucleation is regulated by the recruitment of the ATG14-containing class III phosphatidylinositol 3-kinase (PtdIns3K) complex and the synthesis of PtdIns3P at the phagophore membrane. Subsequently, 2 conjugation systems involving ubiquitin-like (UBL) proteins contribute to the expansion of the phagophore. The first UBL protein system involves the formation of the ATG12–ATG5-ATG16L1 complex. The second conjugation pathway leads to the lipidation of LC3 to phosphatidylethanolamine (PE) resulting in its recruitment to the nascent phagophore structure, a process facilitated by the ATG12–ATG5-ATG16L1 conjugate.

Thus, the conjugation of LC3 to PE is one of the key events of macroautophagy, and it plays a critical role in autophagosome formation. Therefore, one presumed characteristic for the membrane that supports the formation of the phagophore is the capacity to recruit LC3 and promote its efficient conjugation to PE. To identify such a membrane(s), Ge et al. established an elegant in vitro system monitoring the synthesis of LC3-II, the PE conjugated form of LC3. ¹ In a conceptually direct approach, the authors used membranes from Atg5 knockout MEF cells, which are deficient for LC3 lipidation and phagophore formation, as acceptors to which they added recombinant LC3 as well as cytosol from wild-type cells; the latter provides the molecular components necessary for LC3 conjugation.

In a first step, the authors showed that the in vitro lipidation system responds to the pathways regulating autophagy in vivo. Next, sequential fractionation of the Atg5 knockout cell membranes was performed and the in vitro LC3 lipidation activity of the resulting compartments was measured. This approach identified a positive correlation between LC3-II synthesis and the fractions corresponding to the ERGIC, suggesting that these membranes are necessary for, or at least play an important role in, autophagosome biogenesis. This model was supported by use of the drugs H89 and clorofibrate, which deplete the ERGIC and inhibit LC3 in vitro lipidation. The same pharmacological approach, as well as genetic disruption of the ERGIC, further show a dramatic decrease in the formation of LC3 puncta in live cells, thus indicating that ERGIC is crucial for the efficient generation of the autophagosome. Upon induction of autophagy, ATG14 and the PtdIns3P-binding protein ZFYVE1/DFCP1 are rapidly recruited to the phagophore membrane. Depletion of the ERGIC prevents the formation of both ATG14 and ZFYVE1 puncta after starvation, and reduces the membrane recruitment of both proteins. In addition, fluorescence microscopy analyses show colocalization of ATG14 and ZFYVE1 with a marker of the ERGIC after starvation. Because the recruitment of ATG14 and ZFYVE1 occurs upstream of phagophore generation, these results suggest that the ERGIC constitutes a platform for the nucleation of the phagophore membrane.

Previous studies have addressed the topic of phagophore nucleation and expansion, implicating various membranes in providing a scaffold for expansion. For example, upon starvation, ATG14, PtdIns3P, and ZFYVE1 are enriched in specific endoplasmic reticulum (ER) subdomains from which structures called omegasomes emerge. ² The ATG5 complex together with LC3 are later recruited to this structure and it was thus proposed that omegasomes correspond to the platform for phagophore formation. Conversely, work by Hamasaki and colleagues suggested that ATG14 and ZFYVE1 puncta assemble at the mitochondria-associated ER membrane (MAM) under starvation conditions. ³ Disruption of the MAM results in impairment of ATG14 puncta formation and reduction of autophagic activity, suggesting that the ER constitutes the platform for autophagosome formation, and that exchanges between the ER and the mitochondria are necessary for this process. In line with this study, previous work has shown the importance of the MAM in autophagy along with demonstrating the colocalization of ATG5 and LC3 with the mitochondria, and identifying protein from the outer leaflet of the mitochondria on forming autophagosomes. ⁴ It was postulated that the MAM acts in the autophagosome formation process by allowing the transfer of necessary proteins and/or lipids to the nascent phagophore, in particular by providing the PE required for LC3 conjugation. The study by Ge et al., now shows that 1) the MAM is not able to support LC3 lipidation, suggesting that this structure would act later in the pathway of autophagosome formation, and 2) the level of lipidation activity in a particular fraction does not correlate with its level of PE, suggesting that a high abundance of this lipid is not a prerequisite for the nucleation of the phagophore. ¹

Over the years, the search for the membrane source for autophagosome biogenesis has led to the identification of several subcellular compartments potentially involved in that process. Components of the exocyst, a tethering complex that mediates the fusion of post-Golgi vesicles with the plasma membrane associates with nascent autophagosomes and is essential for starvation-induced autophagy. ⁵ Atg9-containing vesicles, which cycle between the trans-Golgi network, post-Golgi, and endosomal compartments, are also crucial for the formation of autophagosomes. ⁶

Recently, the group of Rubinsztein reported that ATG9 also traffics from the plasma membrane to recycling endosomes (see the punctum by Puri et al., in this issue of the journal). ⁷ Previously ATG16L1 was shown to interact with the clathrin heavy chain, an element of the endocytic vesicle coat, and this interaction is critical for the generation of the phagophore. ⁸ It appears that ATG9 and ATG16L1 traffic in different endocytic vesicles to the recycling endosomes where VAM3-dependent heterotypic fusion occurs and that this event is required for autophagosome formation. ⁷ In their fractionation analyses, Ge and coworkers were able to separate the ERGIC from compartments shown to be involved in autophagosome formation: plasma membrane, endosomes, Golgi, mitochondria, or the ATG9-containing vesicles. They showed that of these only the ERGIC is able to support in vitro LC3 lipidation, suggesting that other membrane sources are not necessary for the nucleation step of phagophore biogenesis. ¹ These results support the idea that the first step of phagophore formation occurs at the ERGIC by the recruitment of ATG14 and the lipidation of LC3.

This work does not eliminate the possibility that other membrane compartments contribute to autophagosome formation. It is plausible that, after induction of the precursor structure at the ERGIC, additional membrane sources coalesce at this site thereby increasing the size of the membranes at the phagophore and bringing necessary proteins and lipids allowing for its maturation into an autophagosome. This hypothesis would conciliate the different models previously proposed, and is in line with the idea that multiple membrane sources might be required to sustain autophagosome biogenesis without altering overall cellular homeostasis under starvation conditions.

Altogether, this study supports a model in which a unique membrane, the ERGIC, would respond to stress by providing a platform for phagophore formation. Why is the ERGIC membrane able to constitute such a site? Which features distinguish the ERGIC from other compartments and confer upon it its LC3 lipidation property? Questions also arise concerning the identity of the signal triggering the recruitment of ATG14 at the ERGIC and the mechanism of this recruitment, which will need to be resolved. Additionally, the involvement of other membrane compartments and the timing of their incorporation into the autophagosome biogenesis pathway relative to the ERGIC need to be elucidated.

10.4161/auto.27242

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.