ATG2B is a large conserved autophagy-related lipid-transfer protein that acts at endoplasmic reticulum-phagophore membrane contact sites during autophagosome biogenesis. It binds membranes and forms an ATG2-WIPI/ATG18 complex with WDR45/WIPI4, which promotes ATG2B recruitment to PI3P-containing autophagic membranes and stimulates lipid-transfer activity. Mammalian ATG2 proteins are also associated with lipid droplets and affect lipid droplet morphology and dispersion, but the best-supported core role of ATG2B is membrane tethering and lipid transfer for phagophore expansion.
Definition: The membrane system of a phagophore, comprising its two closely apposed lipid bilayers and the continuous rim that joins them, which expands by lipid acquisition and closes to form the double membrane of an autophagosome.
Justification: GO has autophagosome membrane (GO:0000421) and omegasome membrane (GO:1903349) but no term for the membrane of the phagophore itself, so annotations describing the phagophore membrane have gone to GO:0034045 phagophore assembly site membrane for want of anything better. That term asserts via bounding_layer_of that a membrane bounds the phagophore assembly site, which is a protein condensate with no bounding bilayer (PMID:32025038), and it carries "phagophore" and "isolation membrane" as related synonyms, competing directly with GO:0061908. See GO issue #29437 (recommendation R3). Note the sheet-versus-system question is unresolved upstream - this proposal takes the collective reading; the alternative is "either of the two closely apposed lipid bilayers that form a phagophore".
Parent term: membrane
Definition: A zone of apposition between the endoplasmic reticulum membrane and the phagophore, bridged by ATG2-family lipid transfer proteins together with Atg18/WIPI-family PI3P effectors and the ATG9 scramblase, across which glycerophospholipids are delivered to support phagophore expansion.
Justification: GO models many other contact sites explicitly (GO:0044233, GO:0140268, GO:0120095 vacuole-isolation membrane contact site) but has nothing for the phagophore-ER apposition, which is the best-characterised of the phagophore's contacts. See GO issue #29437 (recommendation R7) and modules/phagophore_organelle_contact_site.yaml.
Parent term: organelle membrane contact site
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0000407 phagophore assembly site | IBA GO_REF:0000033 | ACCEPT | Summary: ATG2B is active at autophagy initiation/elongation sites. Mammalian ATG2 proteins are required at a late step of autophagosome formation, with a direct mechanism of lipid transfer at the ER-phagophore interface. Reason: This cellular-component annotation matches the conserved ATG2/WIPI role at phagophore assembly sites and is supported by mammalian cell evidence for ATG2-dependent autophagosome formation. Supporting Evidence: PMID:22219374 both Atg2A and Atg2B are required for autophagy and that they have redundant and overlapping functions PMID:31721365 direct lipid transfer (LT) from ER to IM for IM expansion |
| GO:0000422 autophagy of mitochondrion | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: ATG2B is part of the core macroautophagy machinery required to build autophagosomal membranes, so it can participate in mitophagy when the macroautophagy machinery is recruited to mitochondria. Reason: The annotation is biologically plausible as a phylogenetic inference for a core autophagy factor, but it should not be interpreted as a mitochondria- selective recognition role for ATG2B. The core function remains phagophore lipid transfer. Supporting Evidence: PMID:22219374 these results suggest that Atg2A/B play an essential role, probably at a late step of autophagosome formation |
| GO:0061908 phagophore | IBA GO_REF:0000033 | ACCEPT | Summary: ATG2B functions on nascent autophagic membranes during phagophore expansion and forms a WIPI/ATG18-associated lipid-transfer module. Reason: The phagophore is the core location for ATG2B lipid-transfer activity in autophagosome biogenesis. Supporting Evidence: PMID:22219374 these results suggest that Atg2A/B play an essential role, probably at a late step of autophagosome formation PMID:31721365 direct lipid transfer (LT) from ER to IM for IM expansion |
| GO:0000425 pexophagy | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: ATG2B can support pexophagy indirectly as part of the general autophagosome-building machinery used by selective autophagy pathways. Reason: This is best retained as a non-core phylogenetic selective-autophagy annotation. The evidence does not show ATG2B as a peroxisome-selective cargo receptor or specificity factor. Supporting Evidence: PMID:22219374 both Atg2A and Atg2B are required for autophagy and that they have redundant and overlapping functions |
| GO:0043495 protein-membrane adaptor activity | IBA GO_REF:0000033 | ACCEPT | Summary: ATG2B binds membranes and WDR45/WIPI4 and functions with this WIPI partner at autophagic membranes, consistent with a protein-membrane adaptor or tethering role in addition to lipid transfer. Reason: Direct biochemical and structural evidence supports ATG2B membrane binding, ATG2B-WDR45 complex formation, and membrane tethering/lipid-transfer activity. The more specific core molecular function is lipid transfer activity, but this term captures a supported tether/adaptor aspect. Supporting Evidence: PMID:28820312 mammalian WDR45/WIPI4 has a stronger binding capacity for mammalian ATG2A or ATG2B than the other 3 WIPIs PMID:31721365 ATG2B possesses the membrane tethering (MT) and LT activity |
| GO:0061723 glycophagy | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: ATG2B may be required for glycophagy insofar as glycophagy uses the core macroautophagy membrane expansion machinery. Reason: Retain as a non-core inferred selective-autophagy annotation. The direct evidence for ATG2B is autophagosome membrane formation, not glycogen cargo recognition. Supporting Evidence: PMID:22219374 both Atg2A and Atg2B are required for autophagy and that they have redundant and overlapping functions |
| GO:0032266 phosphatidylinositol-3-phosphate binding | IBA GO_REF:0000033 | MARK AS OVER ANNOTATED | Summary: The ATG2-WIPI module is recruited to PI3P-positive autophagic membranes, but the direct PI3P-binding effector is the WIPI/ATG18 family partner, not clearly ATG2B itself. Reason: This term overstates the available gene-level evidence for ATG2B. Direct evidence supports liposome binding and WDR45/WIPI4-dependent association with PI3P-containing membranes, but not a specific ATG2B PI3P headgroup binding activity. Supporting Evidence: PMID:28820312 PtdIns3P-binding effectors which can form complexes with proteins in the Atg2 family PMID:28820312 ATG2B and found that it could bind to liposomes independently of PtdIns3P or WDR45 |
| GO:0034727 piecemeal microautophagy of the nucleus | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: ATG2B is a conserved autophagy factor and may support this process where the conserved machinery is used, but human ATG2B is not established as a process-specific PMN factor. Reason: Keep as a non-core phylogenetic inference. The strongest mammalian evidence supports general phagophore/autophagosome biogenesis rather than a human PMN-specific role. Supporting Evidence: PMID:22219374 both Atg2A and Atg2B are required for autophagy and that they have redundant and overlapping functions |
| GO:0061709 reticulophagy | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: ATG2B can support reticulophagy as part of the core autophagy membrane expansion system, especially because ATG2B acts at ER-phagophore contact sites. Reason: Retain as non-core. ATG2B is not a reticulophagy cargo receptor or ER quality-control factor; its direct role is lipid transfer for phagophore expansion. Supporting Evidence: PMID:31721365 direct lipid transfer (LT) from ER to IM for IM expansion |
| GO:0005789 endoplasmic reticulum membrane | IEA GO_REF:0000044 | ACCEPT | Summary: ER membrane localization is consistent with the accepted model in which ATG2B tethers the isolation membrane/phagophore edge to the ER and transfers lipids from ER-derived membranes to the growing phagophore. Reason: Although this GOA row is electronically inferred from UniProt subcellular location, it is consistent with experimental and biochemical evidence for ATG2-dependent ER-to-phagophore lipid transfer. Supporting Evidence: PMID:31721365 direct lipid transfer (LT) from ER to IM for IM expansion |
| GO:0005811 lipid droplet | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: Mammalian ATG2 proteins associate with lipid droplets and affect lipid droplet morphology and dispersion, but this is a secondary cellular context relative to ATG2B's core phagophore lipid-transfer role. Reason: Retain the localization, but mark non-core because the publication notes that the precise lipid-droplet mechanism and directness of the association remain unresolved. Supporting Evidence: PMID:22219374 One novel aspect of the present study is the finding that mammalian Atg2 proteins are present on lipid droplets PMID:22219374 Further experiments will be required to test whether Atg2 proteins directly or indirectly associate with lipid droplets |
| GO:0006914 autophagy | IEA GO_REF:0000002 | MODIFY | Summary: The broad InterPro-derived autophagy annotation is directionally correct but less informative than the experimentally supported role in autophagosome assembly/phagophore expansion. Reason: Modify to GO:0000045 autophagosome assembly. ATG2B is not merely associated with autophagy in general; it is required for formation and closure/expansion of autophagosomal membranes. Proposed replacements: autophagosome assembly Supporting Evidence: PMID:22219374 these results suggest that Atg2A/B play an essential role, probably at a late step of autophagosome formation PMID:31721365 direct lipid transfer (LT) from ER to IM for IM expansion |
| GO:0034045 phagophore assembly site membrane | IEA GO_REF:0000044 | MARK AS OVER ANNOTATED | Summary: The phagophore assembly site membrane is a supported location for ATG2B function and matches its WIPI-associated lipid-transfer role at the growing isolation membrane. Reason: The electronic UniProt location mapping is supported by the experimental ATG2A/B depletion phenotype and by ATG2B membrane tethering/lipid-transfer activity. [2026-08 re-review against GO issue #29437] Held on a defective term with no correct destination in the current ontology, so recorded as an over-annotation with the needed terms proposed rather than left as ACCEPT. GO:0034045 asserts via bounding_layer_of that a membrane bounds the phagophore assembly site, which is a protein condensate with no bounding bilayer (PMID:32025038). ATG2B's tethering and lipid-transfer role places it at the ER-phagophore contact and on the growing phagophore membrane, both proposed under proposed_new_terms. No proposed_replacement_terms are given because no existing term expresses either without loss. Supporting Evidence: PMID:22219374 these results suggest that Atg2A/B play an essential role, probably at a late step of autophagosome formation PMID:31721365 direct lipid transfer (LT) from ER to IM for IM expansion |
| GO:0120009 intermembrane lipid transfer | IEA GO_REF:0000108 | ACCEPT | Summary: ATG2B has membrane tethering and lipid-transfer activity and transfers lipids between membranes during isolation membrane expansion. Reason: This BP annotation follows directly from the experimentally demonstrated lipid-transfer activity and the model of ER-to-isolation-membrane lipid movement. Supporting Evidence: PMID:31721365 ATG2B possesses the membrane tethering (MT) and LT activity PMID:31721365 direct lipid transfer (LT) from ER to IM for IM expansion |
| GO:0005515 protein binding | IPI PMID:20562859 Network organization of the human autophagy system. | MARK AS OVER ANNOTATED | Summary: This interaction record reflects WDR45/WIPI4 association in the human autophagy interaction network, but generic protein binding is not an informative ATG2B molecular function. Reason: The WDR45 interaction is biologically relevant, but the curation should be represented by the ATG2-ATG18 complex/component context and ATG2B lipid transfer/tethering function rather than by GO:0005515. Supporting Evidence: PMID:20562859 association between ATG2A, ATG2B, and WDR45 was unaltered by mTOR inhibition PMID:28820312 mammalian WDR45/WIPI4 has a stronger binding capacity for mammalian ATG2A or ATG2B than the other 3 WIPIs |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | MARK AS OVER ANNOTATED | Summary: This BioPlex interaction row is a high-throughput WDR45 association, but GO:0005515 is too generic for ATG2B. Reason: Retaining a generic protein binding MF would obscure the supported biology: ATG2B forms a WDR45/WIPI4-associated autophagy lipid-transfer complex. The interaction can inform complex membership but should not be treated as a core molecular function by itself. Supporting Evidence: PMID:33961781 These networks model the interactome whose structure encodes protein function, localization, and complex membership. PMID:28820312 mammalian WDR45/WIPI4 has a stronger binding capacity for mammalian ATG2A or ATG2B than the other 3 WIPIs |
| GO:0005515 protein binding | IPI PMID:34524948 Global Proximity Interactome of the Human Macroautophagy Pat... | MARK AS OVER ANNOTATED | Summary: This proximity-interactome annotation supports ATG2B's placement in the human macroautophagy interaction network, but it is not an adequate molecular-function description. Reason: Generic protein binding should not be propagated as ATG2B's function. The informative curation is lipid transfer activity and ATG2-WIPI complex membership. Supporting Evidence: PMID:34524948 Here, we applied BioID to the study of macroautophagy in human cells, generating a proximity interaction map of 39 core macroautophagy proteins. PMID:31721365 ATG2B possesses the membrane tethering (MT) and LT activity |
| GO:0005515 protein binding | IPI PMID:35271311 OpenCell: Endogenous tagging for the cartography of human ce... | MARK AS OVER ANNOTATED | Summary: The OpenCell-derived interaction annotation is useful as interactome context, but GO:0005515 is too broad and uninformative for ATG2B curation. Reason: The interaction should be interpreted conservatively as supporting complex context, not as the gene product's molecular function. ATG2B's curated molecular function is lipid transfer activity. Supporting Evidence: PMID:35271311 well-curated localization and interactome measurements PMID:28820312 We also purified the ATG2B-WDR45 complex and then performed 3-dimensional reconstruction of the complex |
| GO:0005783 endoplasmic reticulum | IDA GO_REF:0000052 | ACCEPT | Summary: ER localization is consistent with ATG2B's role at ER-associated autophagosome biogenesis sites. Reason: The ER is central to ATG2B-mediated lipid transfer to the isolation membrane; this broader ER localization is valid alongside the more specific ER membrane annotation. Supporting Evidence: PMID:31721365 direct lipid transfer (LT) from ER to IM for IM expansion |
| GO:0005789 endoplasmic reticulum membrane | ISS GO_REF:0000024 | ACCEPT | Summary: The ER membrane annotation transferred by sequence similarity is consistent with the experimentally supported ATG2 lipid-transfer model. Reason: ATG2B acts at an ER-to-isolation-membrane interface, so ER membrane localization is a core cellular context. Supporting Evidence: PMID:31721365 direct lipid transfer (LT) from ER to IM for IM expansion |
| GO:0005811 lipid droplet | EXP PMID:22219374 Mammalian Atg2 proteins are essential for autophagosome form... | KEEP AS NON CORE | Summary: ATG2 proteins were observed on lipid droplets, and ATG2A/B depletion altered lipid droplet size and distribution. Reason: This direct localization should be retained, but it is not the main ATG2B core function. The lipid-droplet role is experimentally observed but mechanistically less resolved than ER- phagophore lipid transfer. Supporting Evidence: PMID:22219374 One novel aspect of the present study is the finding that mammalian Atg2 proteins are present on lipid droplets PMID:22219374 These data suggest that mammalian Atg2A and Atg2B function both in autophagosome formation and regulation of lipid droplet volume and distribution. |
| GO:0034045 phagophore assembly site membrane | EXP PMID:22219374 Mammalian Atg2 proteins are essential for autophagosome form... | MARK AS OVER ANNOTATED | Summary: Direct mammalian evidence supports ATG2 localization/function at autophagic membranes and shows defective autophagosome formation after combined ATG2A/B depletion. Reason: This is a core location for ATG2B lipid transfer and membrane tethering during phagophore expansion. [2026-08 re-review against GO issue #29437] Held on a defective term with no correct destination in the current ontology, so recorded as an over-annotation with the needed terms proposed rather than left as ACCEPT. GO:0034045 asserts via bounding_layer_of that a membrane bounds the phagophore assembly site, which is a protein condensate with no bounding bilayer (PMID:32025038). ATG2B's tethering and lipid-transfer role places it at the ER-phagophore contact and on the growing phagophore membrane, both proposed under proposed_new_terms. No proposed_replacement_terms are given because no existing term expresses either without loss. Supporting Evidence: PMID:22219374 these results suggest that Atg2A/B play an essential role, probably at a late step of autophagosome formation |
| GO:0120013 lipid transfer activity | IDA PMID:31721365 Human ATG2B possesses a lipid transfer activity which is acc... | ACCEPT | Summary: Human ATG2B directly possesses lipid-transfer activity, and this activity is promoted by negatively charged membranes and WIPI4. Reason: This is the most informative supported molecular-function annotation for ATG2B and should be treated as the core molecular function. Supporting Evidence: PMID:31721365 ATG2B possesses the membrane tethering (MT) and LT activity PMID:31721365 negatively charged membranes and an Atg18 ortholog WIPI4 |
| GO:0062079 ATG2-ATG18 complex | IDA PMID:28820312 Architecture of the ATG2B-WDR45 complex and an aromatic Y/HF... | NEW | Summary: NEW annotation. ATG2B forms an ATG2-WIPI/ATG18 complex with WDR45/WIPI4, consistent with direct structural and biochemical evidence for complex membership. Reason: This is the conservative way to use the PN projection. The projection is not evidence for a new proteostasis function by itself, but it points to an existing GO complex term that is directly supported by ATG2B-WDR45 structural and biochemical evidence. Supporting Evidence: PMID:28820312 We also purified the ATG2B-WDR45 complex and then performed 3-dimensional reconstruction of the complex PMID:28820312 conserved aromatic H/YF motif in the C terminus of ATG2A and ATG2B that is crucial for complex formation PMID:20562859 association between ATG2A, ATG2B, and WDR45 was unaltered by mTOR inhibition |
Loading supporting content…
Download this section (compressed HTML)Q: Does ATG2B directly associate with lipid droplets at endogenous expression levels, or is the lipid-droplet phenotype secondary to altered ER/phagophore lipid handling?
Suggested experts: Noboru Mizushima, Noriyuki Noda, Li Yu
Q: Is ATG2B's GABARAP/ATG8-family interaction via the conserved ATG2 LIR individually required for ATG2B function at the phagophore (as shown for the ATG2 family), and does it represent a distinct ATG8-binding molecular function separable from WIPI4 binding in mammalian cells?
Suggested experts: David G. McEwan, Noboru Mizushima
Experiment: Use endogenous ATG2B knockout/rescue cells expressing wild-type ATG2B or WDR45-binding motif mutants, then measure ATG2B-WDR45 colocalization, autophagic flux, and phagophore closure under starvation.
Hypothesis: ATG2B-WDR45 complex formation is required for efficient ATG2B recruitment to PI3P-positive phagophore membranes and for autophagosome membrane expansion.
Type: genome editing, rescue, live-cell imaging, and autophagy flux assay
Experiment: In ATG2A/ATG2B double-knockout cells reconstituted with ATG2B, test whether ANKFY1 depletion reduces ATG2B-dependent autophagosome growth and PI3P distribution on phagophores, and assay ANKFY1-ATG2B binding and ATG2B-mediated lipid transfer between PI3P-containing liposomes in vitro.
Hypothesis: ATG2B, like ATG2A, can transfer lipids from PI3P-enriched endosomes (via ANKFY1) in addition to the ER, contributing endosome-derived lipid to phagophore expansion.
Type: knockout/rescue, proximity imaging, in vitro lipid-transfer assay
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)