Mechanisms of autophagosome formation

Abstract

The formation of autophagosomes is a pivotal step in autophagy, a lysosomal degradation system that plays a crucial role in maintaining cellular homeostasis. After autophagy induction, phase separation of the autophagy-related (Atg) 1 complex occurs, facilitating the gathering of Atg proteins and organizes the autophagosome formation site, where the initial isolation membrane (IM)/phagophore is generated. The IM then expands after receiving phospholipids from endomembranes such as the endoplasmic reticulum. This process is driven by the collaboration of lipid transfer (Atg2) and scrambling (Atg9) proteins. The IM assumes a cup shaped morphology and undergoes closure, resulting in the formation of a double membrane-bound autophagosome. The Atg8 lipidation system is hypothesized to be a pivotal factor in this process. This review presents an overview of the current understanding of these processes and discusses the basic mechanisms of autophagosome formation.

1. Introduction

Autophagy is an intracellular degradation system that is conserved among eukaryotes. This process contributes to cellular homeostasis through the degradation of various cytoplasmic materials, including biomolecules such as proteins, nucleic acids, and lipids, as well as organelles such as the endoplasmic reticulum (ER) and mitochondria.¹⁾ This universal degradation ability is made possible through the unique mechanism of autophagy, the de novo formation of autophagosomes (Fig. 1).²⁾ When autophagy is induced, for example by nutrient starvation, isolation membranes (IMs; also known as phagophores) suddenly appear in the cytoplasm, which then expand and seal to form autophagosomes. During this process, cytoplasmic materials are sequestered into the lumens of the autophagosomes that then fuse with lysosomes (vacuoles in yeast and plants); the sequestered materials are then degraded by lysosomal hydrolases. Therefore, in principle, anything that can be sequestered within autophagosomes can be degraded through autophagy.

Fig. 1

(Color online) Schematic drawing of autophagosome formation in yeast.

The genes and proteins responsible for the formation of autophagosomes were initially identified in budding yeast,³⁾ and most of these were later found to be conserved among higher eukaryotes, including mammals.²⁾ In budding yeast, approximately 20 autophagy-related (Atg) proteins are required for the formation of starvation-induced autophagosomes. These proteins have been classified into the following six groups: 1) the autophagy initiating Atg1 complex comprising Atg1, Atg13, Atg17, Atg29, and Atg31; 2) the phosphatidylinositol (PI) 3-kinase complex that produces PI 3-phosphate (PI3P); 3) lipid scramblase Atg9; 4) the Atg2-Atg18 complex that is responsible for lipid transfer; 5) the Atg12-Atg5 conjugation system; and 6) the Atg8 lipidation system comprising Atg8 (a ubiquitin-like protein), Atg4 (processing and deconjugating enzyme), Atg7 (E1), Atg3 (E2), and the Atg12-Atg5-Atg16 complex (E3) that is produced by the Atg12-Atg5 conjugation system.^2),4) Upon starvation, these proteins target the pre-autophagosomal structure (PAS) and function collaboratively to form autophagosomes.⁵⁾

Herein, we summarize current knowledge of the basic mechanisms underlying the formation of starvation-induced autophagosomes, with particular focus on Atg protein-mediated PAS organization and IM generation, expansion, and shaping.

2. Organization of PAS through liquid-liquid phase separation

The basic mechanism of autophagy induction by starvation was understood by the early 2000s. Autophagy induction is directly regulated by target of rapamycin (TOR) kinase.⁶⁾ In growing conditions in yeast, TOR is highly active and directly phosphorylates Atg13 to inhibit the formation of the Atg1 complex and organization of PAS responsible for bulk autophagy.⁷⁾ Upon nutrient starvation, TOR is inhibited and Atg13 is rapidly dephosphorylated, enabling Atg13 to interact with Atg1 and the Atg17-Atg29-Atg31 subcomplex to organize the Atg1 complex, which contributes to PAS assembly and initiates autophagy.^5),7) Later, phosphatases such as PP2A and PP2C were shown to be responsible for Atg13 dephosphorylation upon starvation.^8),9)

The molecular mechanisms involved in the Atg1 complex formation and PAS assembly were then studied through structural biology and in vitro reconstitution approaches. Crystallographic and biochemical analyses revealed that the intrinsically disordered region of Atg13 possesses at least two Atg17-binding regions, which bind to distinct regions in Atg17 independently of each other.^10),11) Notably, owing to distance constraints, the two binding regions of a single Atg13 molecule cannot bind to the same Atg17 molecule or to both of the two Atg17 molecules within the Atg17 homodimer at the same time. Instead, Atg13 crosslinks Atg17 dimers by binding to two Atg17 molecules derived from distinct Atg17 homodimers. As a result, the Atg13-Atg17 interaction can be infinitely repeated in a three-dimensional direction, enabling the Atg1 complex to form a higher-order assemblage.¹⁰⁾ In vitro reconstitution studies revealed that the purified Atg1 complex undergoes liquid-liquid phase separation to form liquid droplets in a manner that depends on the multivalent Atg13-Atg17 interaction (Fig. 2).¹²⁾ Phosphorylation of Atg13 by TOR or point mutations in Atg13 or Atg17 that suppress the Atg13-Atg17 interaction impair phase separation of the Atg1 complex in vitro. Notably, PAS behaves as a liquid droplet, and point mutations that impair phase separation of the Atg1 complex in vitro also impair the formation of PAS in yeast cells.¹²⁾ These observations demonstrated that phase separation of the Atg1 complex also leads to PAS assembly and that TOR-mediated phosphorylation of the Atg1 complex inhibits PAS assembly by blocking phase separation of the Atg1 complex.

Fig. 2

(Color online) Phase separation model of the pre-autophagosomal structure (PAS) assembly.

3. Role of PAS in Atg8 lipidation and initial IM generation

The mechanism through which PAS formation occurs through phase separation is efficient, because it allows Atg factors to quickly gather at one location to form autophagosomes when required. What then is the actual role of PAS as a liquid droplet in autophagosome formation? One possible role is to activate Atg1 kinase, an essential step in autophagy initiation.^7),13) Atg1 kinase activity remains low in growth conditions but changes to high by progressive autophosphorylation upon nutrient starvation. In vitro studies revealed that phase separation of the Atg1 complex accelerates autophosphorylation of Atg1, suggesting that PAS as a liquid droplet promotes autophosphorylation of Atg1 and activates its kinase activity (Fig. 3).¹²⁾ Activation of ULK1/2, the mammalian counterpart of Atg1, is also important for autophagy progression in mammals. The mechanism of ULK1/2 activation, including the involvement of phase separation, is not yet well understood, but the possibility that the p62 body formed by phase separation activates ULK1/2 has been discussed.¹⁴⁾

Fig. 3

(Color online) Multiple roles of the PAS as a liquid droplet in autophagy.

Atg9 is the sole transmembrane protein among the core Atg factors and is embedded in vesicles at the Golgi body.²⁾ In growing conditions, Atg9 vesicles are actively dispersed to the periphery of the cytoplasm, whereas some of the Atg9 vesicle population targets PAS upon starvation. At PAS, Atg9 vesicles are considered as membrane seeds for generating the initial IM.¹⁵⁾ Atg8 is also targeted to PAS upon starvation and is considered to be attached to membranes, such as to Atg9 vesicles, at PAS by E1/E2/E3-mediated conjugation reactions with phosphatidylethanolamine (PE).²⁾ Using an in vitro reconstitution system, we determined that the Atg12-Atg5-Atg16 E3 complex for Atg8 is most efficiently condensed in liquid droplets of the Atg1 complex among the core Atg factors, with the specific Atg12-Atg17 interaction playing an important role. We also showed that the Atg8-PE conjugation reaction with liposomes is significantly promoted by Atg1 complex droplets.¹⁶⁾ In contrast, Atg4, the deconjugase for Atg8, is not condensed at Atg1 complex droplets, and the deconjugation reaction of Atg8-PE by Atg4 is significantly inhibited by Atg1 complex droplets. Notably, the production of Atg8-PE promotes the internalization of liposomes into Atg1 complex droplets. These observations suggest that PAS as a liquid droplet mediates two roles in the Atg8 conjugation system: promoting the conjugation reaction by condensing E3 and inhibiting the deconjugation reaction by eliminating the deconjugase. This proposes an attractive model: PAS pools Atg8-PE-containing Atg9 vesicles for initial IM generation (Fig. 3).¹⁶⁾ Further studies are warranted to clarify the molecular mechanisms underlying initial IM generation and the role of PAS in this process.

4. Autophagosomal membrane expansion by lipid transfer and scrambling proteins

The formation of one autophagosome requires millions of phospholipids as building blocks. Various endomembranes have been proposed to function as membrane sources for autophagosome formation, among which the ER is considered to be the primary membrane source.²⁾ Furthermore, several mechanisms have been proposed for the delivery of phospholipids from the ER to the IM, such as vesicular transport and direct connection models between the ER and the IM.^17)-19) However, recent structural biology and in vitro reconstruction approaches have established a protein-mediated lipid transport model as a novel mechanism for ER to IM phospholipid transfer (Fig. 4A).^20)-23)

Fig. 4

(Color online) (A) Autophagy-related (Atg)2 and Atg9-mediated isolation membrane (IM) expansion model. (B) Architecture of Atg9 (PDB 7D0I).

Atg2 is the largest protein (1,592 amino acids) among the core Atg proteins in budding yeast and possesses a rod-like architecture measuring up to 20 nm with N and C termini at both ends.^24),25) Atg18 is a phosphoinositide-binding protein that binds near the C-terminal edge of the Atg2 rod, and the resultant Atg2-Atg18 complex binds to the expanding edge of the IM through the Atg18-PI3P interaction.^24)-26) In addition, Atg2 binds to the ER, particularly the ER exit site responsible for COPII vesicle formation, through the N-terminal edge of the rod architecture.^27)-29) As a result, the Atg2-Atg18 complex bridges the ER and the IM and generates ER-IM contact (Fig. 1). We performed crystallographic analysis of the N-terminal region of fission yeast Atg2 proteins and revealed that Atg2 possesses a hydrophobic cavity which is used to accommodate phospholipids.²¹⁾ Notably, we demonstrated via in vitro experiments that Atg2 has the function of transferring phospholipids between liposomes.²¹⁾ Consistent with this, several studies by other researchers and ourselves have showed that human ATG2 possesses a lipid transfer activity in vitro,^22),23),30) and cryo-EM analysis of human ATG2 revealed that ATG2 possesses a 15 nm pore along the long axis of the rod architecture.^22),23) AlphaFold also predicted the existence of a 15 nm hydrophobic pore in the proteins belonging to the Atg2 family.³¹⁾ These observations propose a model of protein-mediated lipid transfer from the ER to the IM: the Atg2-Atg18 complex bridges the ER and the IM and directly transfers phospholipids from the ER to the IM through the hydrophobic pore of Atg2.^21)-23) Recently, Vps13, a distant homolog of Atg2, was also shown to play a role in lipid transfer from the ER to the IM in parallel with Atg2, although it is not essential for autophagosome formation.³²⁾

Although Atg2 transfers phospholipids between the outer (cytosolic) leaflets of membranes, it cannot transfer them to or from the inner (luminal) leaflets. In order to accomplish IM expansion, phospholipids must be supplied to both the cytosolic and luminal leaflets of the IM. Atg9 is a transmembrane protein localized at the expanding edge of the IM, together with the Atg2-Atg18 complex.^27),33) We and another group independently revealed via in vitro studies that yeast and human Atg9 possess the activity to transfer phospholipids between the two leaflets of liposomes.^34),35) This lipid scrambling activity does not require ATP, suggesting that Atg9 family proteins are members of the lipid scramblase family. Furthermore, we and other groups performed cryo-EM analyses of Atg9 family proteins and unveiled their unique architecture composed of a domain-swapped homo-trimer with two distinct types of pores, one of which is located at each protomer and laterally opens to the membrane, whereas the other is located at the center of the trimer and vertically penetrates the membrane (Fig. 4B).^34)-37) Mutational analyses revealed that both lateral and vertical pores are important for the lipid scramblase activity of Atg9-family proteins and for autophagosome formation, confirming that lipid scramblase activity is involved in autophagosome formation.^34),35) In combination with Atg2 observations, an attractive model of membrane expansion has been proposed: Atg2 bridges the ER and the IM and transfers phospholipids from the cytosolic leaflet of the ER to the cytosolic leaflet of the IM, where Atg9 scrambles phospholipids and distributes them to both leaflets of the IM, thereby enabling IM expansion (Fig. 4A).^{31),34),35),38),39)}

Compared with direct connection and vesicular transport models, the Atg2- and Atg9-mediated lipid transport model is attractive in that it can account for the unique membrane features of autophagosomes, which lack most transmembrane proteins.^40),41) However, this model has certain limitations, such as a lack of understanding of the mechanism of unidirectional lipid transport from the ER to the IM and the mechanism of collaboration between Atg2 and Atg9. Recent cryo-EM analyses showed that human ATG2 and ATG9 form a 1:3 complex and that the exit of the lipid transferring pore of the C-terminal region of ATG2 faces the entrance of the ATG9 pores responsible for lipid scrambling, proposing a direct lipid flow from ATG2 to ATG9; however, two independent works proposed mutually distinct ATG2-ATG9 connection models.^42),43) Intriguingly, cryo-EM analysis also detected a quite distinct mode of interaction between ATG2 and ATG9, in which ATG9 binds to the entrance of the lipid transferring pore located at the N-terminal region of ATG2.⁴³⁾ In mammals, it has been reported that ATG9 vesicles transiently come to the site of autophagosome formation and rapidly disengage.⁴⁴⁾ It is possible that ATG9 vesicles may act as a source of lipid supply via ATG2, in which case it would be more reasonable for ATG9 to bind to the N-terminal region of ATG2. Further studies are required to confirm the IM expansion mechanism mediated by the collaboration of Atg2 and Atg9 family proteins.

5. Autophagosomal membrane shaping by ubiquitin-like conjugation machinery

Atg8 is a ubiquitin-like protein conjugated with PE via reactions mediated by Atg7 (E1), Atg3 (E2), and the Atg12-Atg5-Atg16 complex (E3) (Fig. 5A).^2),4),45) Atg8-PE is tightly attached to the membranes of all types of autophagosomes, from initial IMs to completed autophagosomes, and it is considered to play crucial roles in autophagosome formation. Atg8-PE shows various membrane activities in vitro, such as tethering, hemi- and full-fusion, and perturbation.^45)-49) However, linking each Atg8-PE membrane activity observed in vitro to autophagic processes in vivo is difficult. Here, we describe the roles of the Atg8 lipidation system, rather than Atg8-PE alone, in IM shaping proposed by in vitro reconstitution approaches.^50)-52)

Fig. 5

(A) Atg8 lipidation system. (B) IM shaping model by Atg8-PE/LC3B-PE and E3 (left) and by Atg8-PE and E1/E2/E3 (right).

Using a spherical giant unilamellar vesicle (GUV) as a Atg8 lipidation site, Dr.Wollert’s group proposed a positive feedback loop model for Atg8 lipidation: The Atg12-Atg5-Atg16 complex, in trace amounts, first attaches to the GUV membrane, where it catalyzes the lipidation of Atg8, and the resultant Atg8-PE recruits more Atg12-Atg5-Atg16 complexes, further increasing Atg8-PE production.⁵¹⁾ Accumulation of sufficient amounts of Atg8-PE and Atg12-Atg5-Atg16 complex promotes the formation of a meshwork structure on membranes, for which Atg16-Atg16 (each from distinct dimeric Atg12-Atg5-Atg16 complexes) and Atg8-Atg12 interactions play crucial roles. It was proposed that the formation of the Atg8-PE and Atg12-Atg5-Atg16 meshwork on the convex surface of the IM promotes its shaping during autophagosome formation (Fig. 5B, top left).⁵¹⁾ Dr. Wollert’s group also performed an LC3B (a mammalian Atg8 homolog) lipidation reaction on spherical GUVs or planar bilayer membranes and showed that LC3B-PE and the ATG12-ATG5-ATG16L1 complex induce the membranes to form a cup shape (Fig. 5B, bottom left).⁵²⁾ Notably, targeting of an ATG16L1-binding protein, WIPI2 (yeast Atg18 homolog), to the plasma membrane in mammalian cells promoted targeting of ATG16L1 and LC3B-PE to the plasma membrane and generation of a cup-shaped membrane from the plasma membrane. They found that both ATG16L1 and LC3B-PE are essential for formations of the cup-shaped membrane and proposed that these factors regulate IM shaping.⁵²⁾

We performed an Atg8 lipidation reaction using non-spherical GUVs (because IMs are not spherical) and detected a remarkable morphological change in GUVs upon Atg8-PE production by live imaging: a prolate shape was turned into a spherical shape with in-bud.⁵⁰⁾ This morphological change is reminiscent of that observed during autophagosome formation, suggesting that the Atg8 lipidation system is responsible for IM shaping. Consistent with the above-described report,⁵¹⁾ a meshwork structure was observed on membranes upon Atg8 lipidation. However, NMR studies failed to detect a direct Atg8-Atg12 interaction. Furthermore, chemical linking of Atg8 to membranes does not cause a morphological change in prolate GUVs even after the addition of the Atg12-Atg5-Atg16 complex. Notably, further addition of Atg7 and Atg3 induces a morphological change into a sphere with an inwards bud.⁵⁰⁾ These observations suggest that the meshwork structure on the membrane is constructed by Atg8-PE and all E1/E2/E3 enzymes (Fig. 5B, right). NMR studies also detected multiple weak interactions between Atg8, E1, E2, and E3, which may be a driver of flexible meshwork formation on membranes.⁵⁰⁾ These observations suggest that the E1, E2, and E3 enzymes for Atg8 are directly involved in IM shaping using non-enzymatic activities. This non-enzymatic role of Atg7, Atg3, and Atg12-Atg5-Atg16 may be accomplished by their unique structures, which are remarkably different from those of canonical E1, E2, and E3, respectively.^4),45)

6. Concluding remarks

The introduction of in vitro reconstitution systems and structural biology has dramatically advanced mechanistic studies of autophagy and improved our understanding of autophagosome formation mechanisms. However, autophagosome formation still remains a major enigma, particularly the de novo generation of the initial IM. Furthermore, the mechanism of unidirectional lipid transport from the ER to the IM also lacks a clear understanding. We hope that an integrated analysis of cell biology, in vitro reconstitution, and structural biology studies will help unravel these critical issues and provide a complete picture of autophagosome formation mechanisms in the near future.

Acknowledgments

This work was supported in part by JSPS KAKENHI Grant Number JP23H02429, JP23H04923, JP23K27122 (to Y.F.), JP23K20044, JP19H05707, JP24H00060 (to N.N.N.), PRIME, Japan Agency for Medical Research and Development Grant number JP20gm6410009 (to Y.F.), CREST, Japan Science and Technology Agency Grant number JPMJCR20E3 (to N.N.N.), and grants from the Takeda Science Foundation (to Y.F., N.N.N.).

Notes

Edited by Yoshinori OHSUMI, M.J.A.

Correspondence should be addressed to: N.N. Noda, Institute for Genetic Medicine, Hokkaido University, Kita 15, Nishi 7, Kita-ku, Sapporo, Hokkaido 060-0815, Japan (e-mail: [e-mail]).

References

• 1)  Morishita, H. and Mizushima, N. (2019) Diverse cellular roles of autophagy. Annu. Rev. Cell. Dev. Biol. 35, 453-475.
• 2)  Nakatogawa, H. (2020) Mechanisms governing autophagosome biogenesis. Nat. Rev. Mol. Cell Biol. 21, 439-458.
• 3)  Tsukada, M. and Ohsumi, Y. (1993) Isolation and characterization of autophagy-defective mutants of Saccharomyces cerevisiae. FEBS Lett. 333, 169-174.
• 4)  Noda, N.N. and Inagaki, F. (2015) Mechanisms of autophagy. Annu. Rev. Biophys. 44, 101-122.
• 5)  Suzuki, K., Kirisako, T., Kamada, Y., Mizushima, N., Noda, T. and Ohsumi, Y. (2001) The pre-autophagosomal structure organized by concerted functions of APG genes is essential for autophagosome formation. EMBO J. 20, 5971-5981.
• 6)  Noda, T. and Ohsumi, Y. (1998) Tor, a phosphatidylinositol kinase homologue, controls autophagy in yeast. J. Biol. Chem. 273, 3963-3966.
• 7)  Kamada, Y., Funakoshi, T., Shintani, T., Nagano, K., Ohsumi, M. and Ohsumi, Y. (2000) Tor-mediated induction of autophagy via an Apg1 protein kinase complex. J. Cell Biol. 150, 1507-1513.
• 8)  Memisoglu, G., Eapen, V.V., Yang, Y., Klionsky, D.J. and Haber, J.E. (2019) PP2C phosphatases promote autophagy by dephosphorylation of the Atg1 complex. Proc. Natl. Acad. Sci. U.S.A. 116, 1613-1620.
• 9)  Yeasmin, A.M., Waliullah, T.M., Kondo, A., Kaneko, A., Koike, N. and Ushimaru, T. (2016) Orchestrated action of PP2A antagonizes Atg13 phosphorylation and promotes autophagy after the inactivation of TORC1. PLoS One 11, e0166636.
• 10)  Yamamoto, H., Fujioka, Y., Suzuki, S.W., Noshiro, D., Suzuki, H., Kondo-Kakuta, C. et al. (2016) The intrinsically disordered protein Atg13 mediates supramolecular assembly of autophagy initiation complexes. Dev. Cell 38, 86-99.
• 11)  Fujioka, Y., Suzuki, S.W., Yamamoto, H., Kondo-Kakuta, C., Kimura, Y., Hirano, H. et al. (2014) Structural basis of starvation-induced assembly of the autophagy initiation complex. Nat. Struct. Mol. Biol. 21, 513-521.
• 12)  Fujioka, Y., Alam, J.M., Noshiro, D., Mouri, K., Ando, T., Okada, Y. et al. (2020) Phase separation organizes the site of autophagosome formation. Nature 578, 301-305.
• 13)  Noda, N.N. and Fujioka, Y. (2015) Atg1 family kinases in autophagy initiation. Cell. Mol. Life Sci. 72, 3083-3096.
• 14)  Turco, E., Witt, M., Abert, C., Bock-Bierbaum, T., Su, M.Y., Trapannone, R. et al. (2019) FIP200 claw domain binding to p62 promotes autophagosome formation at ubiquitin condensates. Mol. Cell 74, 330-346.e311.
• 15)  Yamamoto, H., Kakuta, S., Watanabe, T.M., Kitamura, A., Sekito, T., Kondo-Kakuta, C. et al. (2012) Atg9 vesicles are an important membrane source during early steps of autophagosome formation. J. Cell Biol. 198, 219-233.
• 16)  Fujioka, Y., Tsuji, T., Kotani, T., Kumeta, H., Kakuta, C., Fujimoto, T. et al. (2024) Phase separation promotes Atg8 lipidation for autophagy progression. bioRxiv 2024.08.29.610189.
• 17)  Hayashi-Nishino, M., Fujita, N., Noda, T., Yamaguchi, A., Yoshimori, T. and Yamamoto, A. (2009) A subdomain of the endoplasmic reticulum forms a cradle for autophagosome formation. Nat. Cell Biol. 11, 1433-1437.
• 18)  Shima, T., Kirisako, H. and Nakatogawa, H. (2019) COPII vesicles contribute to autophagosomal membranes. J. Cell Biol. 218, 1503-1510.
• 19)  Wang, J., Davis, S., Menon, S., Zhang, J., Ding, J., Cervantes, S. et al. (2015) Ypt1/Rab1 regulates Hrr25/CK1δ kinase activity in ER-Golgi traffic and macroautophagy. J. Cell Biol. 210, 273-285.
• 20)  Osawa, T. and Noda, N.N. (2019) Atg2: A novel phospholipid transfer protein that mediates de novo autophagosome biogenesis. Protein Sci. 28, 1005-1012.
• 21)  Osawa, T., Kotani, T., Kawaoka, T., Hirata, E., Suzuki, K., Nakatogawa, H. et al. (2019) Atg2 mediates direct lipid transfer between membranes for autophagosome formation. Nat. Struct. Mol. Biol. 26, 281-288.
• 22)  Valverde, D.P., Yu, S., Boggavarapu, V., Kumar, N., Lees, J.A., Walz, T. et al. (2019) ATG2 transports lipids to promote autophagosome biogenesis. J. Cell Biol. 218, 1787-1798.
• 23)  Maeda, S., Otomo, C. and Otomo, T. (2019) The autophagic membrane tether ATG2A transfers lipids between membranes. Elife 8, e45777.
• 24)  Tamura, N., Nishimura, T., Sakamaki, Y., Koyama-Honda, I., Yamamoto, H. and Mizushima, N. (2017) Differential requirement for ATG2A domains for localization to autophagic membranes and lipid droplets. FEBS Lett. 591, 3819-3830.
• 25)  Zheng, J.X., Li, Y., Ding, Y.H., Liu, J.J., Zhang, M.J., Dong, M.Q. et al. (2017) Architecture of the ATG2B-WDR45 complex and an aromatic Y/HF motif crucial for complex formation. Autophagy 13, 1870-1883.
• 26)  Obara, K., Sekito, T., Niimi, K. and Ohsumi, Y. (2008) The Atg18-Atg2 complex is recruited to autophagic membranes via phosphatidylinositol 3-phosphate and exerts an essential function. J. Biol. Chem. 283, 23972-23980.
• 27)  Suzuki, K., Akioka, M., Kondo-Kakuta, C., Yamamoto, H. and Ohsumi, Y. (2013) Fine mapping of autophagy-related proteins during autophagosome formation in Saccharomyces cerevisiae. J. Cell Sci. 126, 2534-2544.
• 28)  Graef, M., Friedman, J.R., Graham, C., Babu, M. and Nunnari, J. (2013) ER exit sites are physical and functional core autophagosome biogenesis components. Mol. Biol. Cell 24, 2918-2931.
• 29)  Kotani, T., Kirisako, H., Koizumi, M., Ohsumi, Y. and Nakatogawa, H. (2018) The Atg2-Atg18 complex tethers pre-autophagosomal membranes to the endoplasmic reticulum for autophagosome formation. Proc. Natl. Acad. Sci. U.S.A. 115, 10363-10368.
• 30)  Osawa, T., Ishii, Y. and Noda, N.N. (2020) Human ATG2B possesses a lipid transfer activity which is accelerated by negatively charged lipids and WIPI4. Genes Cells 25, 65-70.
• 31)  Osawa, T., Matoba, K. and Noda, N.N. (2022) Lipid transport from endoplasmic reticulum to autophagic membranes. Cold Spring Harb. Perspect. Biol. 14, a041254.
• 32)  Dabrowski, R., Tulli, S. and Graef, M. (2023) Parallel phospholipid transfer by Vps13 and Atg2 determines autophagosome biogenesis dynamics. J. Cell Biol. 222, e202211039.
• 33)  Gomez-Sanchez, R., Rose, J., Guimaraes, R., Mari, M., Papinski, D., Rieter, E. et al. (2018) Atg9 establishes Atg2-dependent contact sites between the endoplasmic reticulum and phagophores. J. Cell Biol. 217, 2743-2763.
• 34)  Matoba, K., Kotani, T., Tsutsumi, A., Tsuji, T., Mori, T., Noshiro, D. et al. (2020) Atg9 is a lipid scramblase that mediates autophagosomal membrane expansion. Nat. Struct. Mol. Biol. 27, 1185-1193.
• 35)  Maeda, S., Yamamoto, H., Kinch, L.N., Garza, C.M., Takahashi, S., Otomo, C. et al. (2020) Structure, lipid scrambling activity and role in autophagosome formation of ATG9A. Nat. Struct. Mol. Biol. 27, 1194-1201.
• 36)  Lai, L.T.F., Yu, C., Wong, J.S.K., Lo, H.S., Benlekbir, S., Jiang, L. et al. (2020) Subnanometer resolution cryo-EM structure of Arabidopsis thaliana ATG9. Autophagy 16, 575-583.
• 37)  Guardia, C.M., Tan, X.F., Lian, T., Rana, M.S., Zhou, W., Christenson, E.T. et al. (2020) Structure of human ATG9A, the only transmembrane protein of the core autophagy machinery. Cell Rep. 31, 107837.
• 38)  Ghanbarpour, A., Valverde, D.P., Melia, T.J. and Reinisch, K.M. (2021) A model for a partnership of lipid transfer proteins and scramblases in membrane expansion and organelle biogenesis. Proc. Natl. Acad. Sci. U.S.A. 118, e2101562118.
• 39)  Noda, N.N. (2021) Atg2 and Atg9: Intermembrane and interleaflet lipid transporters driving autophagy. Biochim. Biophys. Acta Mol. Cell Biol. Lipids 1866, 158956.
• 40)  Baba, M., Osumi, M. and Ohsumi, Y. (1995) Analysis of the membrane structures involved in autophagy in yeast by freeze-replica method. Cell Struct. Funct. 20, 465-471.
• 41)  Fengsrud, M., Erichsen, E.S., Berg, T.O., Raiborg, C. and Seglen, P.O. (2000) Ultrastructural characterization of the delimiting membranes of isolated autophagosomes and amphisomes by freeze-fracture electron microscopy. Eur. J. Cell Biol. 79, 871-882.
• 42)  van Vliet, A.R., Chiduza, G.N., Maslen, S.L., Pye, V.E., Joshi, D., De Tito, S. et al. (2022) ATG9A and ATG2A form a heteromeric complex essential for autophagosome formation. Mol. Cell 82, 4324-4339.e4328.
• 43)  Wang, Y., Dahmane, S., Ti, R., Mai, X., Zhu, L., Carlson, L.A. et al. (2024) Structural basis for lipid transfer by the ATG2A-ATG9A complex. Nat. Struct. Mol. Biol. doi: 10.1038/s41594-024-01376-6.
• 44)  Orsi, A., Razi, M., Dooley, H.C., Robinson, D., Weston, A.E., Collinson, L.M. et al. (2012) Dynamic and transient interactions of Atg9 with autophagosomes, but not membrane integration, are required for autophagy. Mol. Biol. Cell 23, 1860-1873.
• 45)  Noda, N.N. (2023) Structural biology of the Atg8 and Atg12 conjugation systems. Autophagy Rep. 2, 2277582.
• 46)  Maruyama, T., Alam, J.M., Fukuda, T., Kageyama, S., Kirisako, H., Ishii, Y. et al. (2021) Membrane perturbation by lipidated Atg8 underlies autophagosome biogenesis. Nat. Struct. Mol. Biol. 28, 583-593.
• 47)  Nakatogawa, H., Ichimura, Y. and Ohsumi, Y. (2007) Atg8, a ubiquitin-like protein required for autophagosome formation, mediates membrane tethering and hemifusion. Cell 130, 165-178.
• 48)  Weidberg, H., Shpilka, T., Shvets, E., Abada, A., Shimron, F. and Elazar, Z. (2011) LC3 and GATE-16 N termini mediate membrane fusion processes required for autophagosome biogenesis. Dev. Cell 20, 444-454.
• 49)  Knorr, R.L., Nakatogawa, H., Ohsumi, Y., Lipowsky, R., Baumgart, T. and Dimova, R. (2014) Membrane morphology is actively transformed by covalent binding of the protein Atg8 to PE-lipids. PLoS One 9, e115357.
• 50)  Alam, J.M., Maruyama, T., Noshiro, D., Kakuta, C., Kotani, T., Nakatogawa, H. et al. (2024) Complete set of the Atg8-E1-E2-E3 conjugation machinery forms an interaction web that mediates membrane shaping. Nat. Struct. Mol. Biol. 31, 170-178.
• 51)  Kaufmann, A., Beier, V., Franquelim, H.G. and Wollert, T. (2014) Molecular mechanism of autophagic membrane-scaffold assembly and disassembly. Cell 156, 469-481.
• 52)  Mohan, J., Moparthi, S.B., Girard-Blanc, C., Campisi, D., Blanchard, S., Nugues, C. et al. (2024) ATG16L1 induces the formation of phagophore-like membrane cups. Nat. Struct. Mol. Biol. 31, 1448-1459.

Non-standard abbreviation list

Atg

autophagy-related

ER

endoplasmic reticulum

GUV

giant unilamellar vesicle

IM

isolation membrane

PAS

pre-autophagosomal structure

PE

phosphatidylethanolamine

PI

phosphatidylinositol

PI3P

phosphatidylinositol 3-phosphate

TOR

target of rapamycin

Profile

Yuko Fujioka was born in Hokkaido Prefecture, Japan, in 1976 and graduated from Shizuoka University. She pursued her graduate studies at the Graduate School of Pharmaceutical Sciences, Hokkaido University, earning her PhD in 2009 under the mentorship of Professor Fuyuhiko Inagaki. Following a postdoctoral fellowship, she joined the Institute of Microbial Chemistry, Research Institute for Microbial Chemistry as a researcher in 2011, advancing to the position of senior researcher in 2017. In 2022, she transitioned to the Institute for Genetic Medicine at Hokkaido University, where she has served as both Assistant Professor and Associate Professor. With over two decades of dedicated research into the structural biology and biochemistry of autophagy, she has made significant contributions to the field, including elucidation of the mechanism of autophagy initiation by liquid-liquid phase separation. In recent years, she has been focusing on the study of liquid-liquid phase separation in autophagy and other cellular events.

Nobuo N. Noda was born in Saitama Prefecture, Japan, in 1973 and graduated from the University of Tokyo School of Pharmaceutical Sciences in 1996. He majored in structural biology using X-ray crystallography and received his PhD degree in 2001. He worked as a postdoctoral fellow, assistant professor, and then lecturer at the Graduate School of Pharmaceutical Sciences, Hokkaido University between 2001 and 2011. He became a principal investigator at the Institute of Microbial Chemistry in 2011 and a professor at the Institute for Genetic Medicine, Hokkaido University in 2022. He performed pioneering work on the molecular mechanisms of autophagy, including the discovery of a lipid transfer protein and lipid scramblase mediating autophagosomal membrane formation. For his achievement, he received the Young Investigator Award from the Japanese Biochemistry Society and the Inoue Prize for Science.