Atg2/TRAPPIII-Ypt1 axis: deciphering the phagophore-ERES connection

ABSTRACT

De novo generation of membrane contact sites (MCSs) between the nascent phagophore and the endoplasmic reticulum (ER), particularly the ER exit sites (ERES), are crucial for autophagy as they provide the lipids necessary for the phagophore expansion into an autophagosome. Our recent study provides insights into the mechanism involved in the formation of phagophore-ERES MCSs and uncovers how this event synchronizes the factors involved in phagophore expansion. We revealed that the TRAPPIII complex, the guanine nucleotide exchange factor of the Rab GTPase Ypt1, and the lipid transfer protein Atg2 participate in the phagophore-ERES association. We also show that establishment of phagophore-ERES MCSs leads to TRAPPIII activation and subsequent Ypt1 recruitment onto the phagophore. The presence of active Ypt1 on the growing phagophore enhances local biosynthesis of phosphatidylinositol-3-phosphate (PtdIns3P), triggering the recruitment of the PtdIns3P-effectors Atg18 and Atg21, which play a central role in phagophore expansion. These findings suggest that generation of phagophore-ERES MCSs is one of the signals initiating phagophore expansion.

Abbreviations: Atg, autophagy related; ER, endoplasmic reticulum; ERES, ER exit sites; GEF, guanine nucleotide exchange factor; MCS, membrane contact site; PAS, phagophore assembly site; PtdIns3P, phosphatidylinositol-3-phosphate; PtdIns3K, phosphatidylinositol 3-kinase; SNARE, soluble NSF attachment protein receptor; TOR, Target of Rapamycin; WIPI, WD-repeat domain, phosphoinositide interacting.

Membrane contact sites (MCSs) are critical hubs for inter-organellar communication. At these locations, specific proteins and protein complexes facilitate non-vesicular transfer of lipids, calcium ions, and other small molecules, enabling direct and rapid exchange of specific cellular components between organelles. The MCSs between the phagophore and the endoplasmic reticulum (ER) are particularly relevant, as they supply lipids necessary for phagophore expansion. In Saccharomyces cerevisiae, part of the phagophore extremity is connected to ER exit sites (ERES), which are specialized subdomains of the ER. Our previous work has shown that the interaction between Atg2 and Atg9 is required to establish these phagophore-ERES MCSs. However, ATG2 depletion does not abolish the association between the phagophore and the ER/ERES, suggesting the presence of at least one other tethering factor.

Our recent study has provided some insights into how the phagophore-ERES MCSs are assembled and unraveled an undocumented role in organizing the molecular machinery required for phagophore expansion [1]. We showed that the phagophore assembly site (PAS), the site where the Atg machinery converges to first generate a phagophore and then expand it into an autophagosome, is tethered with both the ERES and vacuolar membrane as soon as it is formed, until autophagosome completion, suggesting that the association with these two organelles acts as an organizational hub during autophagosome formation. During bulk autophagy, the early PAS is a liquid-like biomolecular condensate mainly formed by the Atg1 complex that also recruits Atg9-positive vesicles, which act as membrane precursors and contribute to the phagophore nucleation. Mostly based on in vitro experiments, the model was that Atg9/ATG9A (yeast/mammalian)-positive vesicles establish MCSs with the ER via Atg2/ATG2 proteins. In our study, we showed in vivo that the de novo generation of phagophore-ERES MCSs is not only mediated by Atg2 but also requires the TRAPPIII complex, the guanine nucleotide exchange factor (GEF) of the Rab GTPase Ypt1. We confirmed that, although TRAPPIII and Atg2 do not depend on each other for their recruitment to the phagophore, Atg2 association with the ERES involves TRAPPIII and vice versa. TRAPPIII mediates the association of COPII-coated vesicles with the cis-Golgi, via activation of Ypt1 and subsequent recruitment of tethering factors (such as Uso1/p115), which in turn regulate the engagement of the soluble NSF attachment protein receptor (SNARE) fusion machinery. Analogously, we found that TRAPPIII localizes to the phagophore edge, similarly to Atg2, and promotes the association between the phagophore and the ERES. TRAPPIII appears to require Atg2 at the phagophore-ER MCSs and Atg2 binding to Atg9, not for its recruitment, but for triggering its Ypt1-GEF activity. Conversely, TRAPPIII and Ypt1 are not involved in the association of Atg9-positive vesicles with the PAS or for recruitment of the Atg1 kinase complex or the class III phosphatidylinositol 3-kinase (PtdIns3K) complex I, which together with Atg9-positive vesicles, play a central role in phagophore nucleation. However, Ypt1 is critical to enhance PtdIns3K complex I activity, which leads to higher PtdIns3P levels at the PAS/phagophore. This is consistent with in vitro studies that have shown that active GTP-bound RAB1A, one of the mammalian homologs of Ypt1, binds and stimulates PtdIns3P synthesis by PtdIns3K complex I. Our in vivo results confirm the positive regulation of PtdIns3K complex I by Ypt1 and mechanistically integrate this regulation into the process of autophagosome formation by showing that higher PtdIns3P levels leads to the recruitment of effectors from the WIPI (WD repeat domain, phosphoinositide interacting) protein family, such as Atg18 and Atg21. This latter event appears to be central in initiating phagophore expansion because the presence of Atg18 and Atg21 induces Atg2 lipid transfer activity and activation of the Atg ubiquitin-like conjugation systems, respectively (Figure 1).

Figure 1.

Model for how phagophore-ERES MCS regulates phagophore expansion. Atg9-positive vesicles, which also carry the TRAPPIII complex, are recruited to the PAS upon autophagy induction. The presence of Atg9 and minimal PtdIns3P levels produced by the PtdIns3K complex I (PtdIns3k I) trigger Atg2 recruitment to the PAS via coincidence binding. Both TRAPPIII and Atg2 interact with one or more components of the ERES, promoting the establishment of the phagophore-ERES MCSs. Once a phagophore-ERES MCS is assembled, Atg2 stimulates the GEF activity of the TRAPPIII complex (highlighted in yellow), which triggers the recruitment of the active GTP-bound form of Ypt1 onto the phagophore. Active Ypt1 enhances the PtdIns3K complex I activity (highlighted in light blue), leading to increased PtdIns3P levels at the phagophore, which promote the recruitment of PtdIns3P-effectors such as Atg18 and Atg21. Whereas Atg18 stimulates the Atg2-mediated lipid transfer from the ER to the phagophore, Atg21 activates the ubiquitin-like conjugation systems. These two players are central for phagophore expansion.

The endomembrane system is a complex network of organelles in eukaryotic cells that work together to generate, modify and transport proteins and lipids. This system includes the ER, Golgi apparatus, endosomes and other organelles, all connected through vesicular trafficking pathways and MCSs. Ypt/RAB GTPases mark each of these organelles, reflecting their broad cellular relevance as organelle signature proteins. Indeed, they act as master regulators of membrane trafficking and organelle identity, orchestrating events such membrane fusion, membrane fission, intracellular organelle positioning and organelle functions, which together define cellular compartmentalization.

Our findings also suggest that the cellular roles of MCSs extend beyond simple structural organization and metabolite transport, and they may trigger signaling cascades that would lead to functional changes such as the rewiring of metabolic pathways and/or membrane remodeling. In particular, establishment of the phagophore-ERES MCS is a prerequisite for Ypt1 activation, a molecular switch that promotes phagophore expansion. A few Ypt/RAB GTPases localize to specific MCSs, where they are thought to primarily have an organizational function. Examples are RAB18, which is present at the ER-lipid droplet MCSs and participates in lipid droplet biogenesis, and RAB5 and RAB7, which are found at the ER-endosome and mitochondria-endosome MCSs, respectively, and regulate multiple functions of early and late endosomes. Thus, a speculative idea is that these RAB GTPases as well as others, like Ypt1, act as signaling hubs to coordinate MCS establishment with organelle function.

Ypt1 and its mammalian RAB1 homologs participate in several key pathways, including ER-Golgi transport, endosome-Golgi trafficking and autophagy. Although the involvement of TRAPPIII and Ypt1/RAB1 in autophagy has been previously shown by several studies, their role during autophagosome formation was far from being understood. Our results provide at least one mechanism for how Ypt1 is engaged at the ERES interface toward autophagy instead of ER-Golgi transport. However, based on reported functional interactions with Atg1, Atg11, Atg17, Atg23 and the target of rapamycin complex 1/TORC1, one could envision that Ypt1 is recruited to autophagosomal intermediates through other mechanisms and/or that this small GTPase may regulate other components of the ATG machinery.

In summary, our work unraveled a crucial regulation step in autophagosome biogenesis that occurs at the phagophore-ERES MCSs. Our results suggest that through a regulatory signal centered around Ypt1, cells ensure that phagophore expansion begins only when the major lipid source required for this process is secured.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Funding Statement

F.R. is supported by Novo Nordisk Foundation (0066384) and Lundbeck Foundation (R383-2022–180) grants. C.U. is supported by DFG grants (37802001 and 516911046). J.C.F. is supported by a NIH/NIGMS grant (R35GM136258). The authors acknowledge financial support through Gravitation grant “FLOW” (024.006.036) from the Dutch Ministry of Education, Culture, and Science (OCW) (R. G.-S.).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.