| Aspect | Key finding | Evidence type (review/primary) | System/model | Key experimental readout or statistic | Citation (first author year) | URL/DOI | Notes on ATG2A vs ATG2B specificity |
|---|---|---|---|---|---|---|---|
| Identity/family | Human ATG2B is the ATG2-family paralog of ATG2A; both are very large rod-like autophagy factors with conserved N-terminal chorein/VPS13-like region, ATG2_CAD, ATG_C, a C-terminal localization region, and a conserved LIR near the WIPI4-interaction region. | Review | Mammalian/yeast comparative literature synthesis | ATG2A is ~44.5% identical to ATG2B; single depletion of either paralog does not abolish autophagy. (pqac-00000003, pqac-00000013) | Duarte 2023 | https://doi.org/10.1177/25152564231183898 | Strong family-level evidence; most mechanistic papers test ATG2A directly and infer partial redundancy with ATG2B. |
| Function | ATG2 proteins act as lipid transfer proteins and membrane tethers that support phagophore expansion at ER–phagophore contact sites. | Review | Mammalian autophagy field overview | ATG2 transfers phospholipids from ER to phagophore; ATG9A scramblase activity complements transfer by equilibrating leaflets. (pqac-00000024, pqac-00000009) | McEwan 2022 | https://doi.org/10.1111/febs.16280 | Function generally assigned to ATG2A/ATG2B together in mammals; direct in vitro assays are mostly with ATG2A. |
| Function | ATG2-mediated lipid transfer is sufficient to restore autophagy in ATG2A/ATG2B double-knockout cells when an active N-terminal ATG2A mini-fragment is expressed. | Primary | Human ATG2 DKO cells with rescue constructs | Mini-ATG2A(1–345) rescued autophagy; near-complete rescue required ~10× endogenous expression; lipid-transfer-defective mutants failed to rescue. (pqac-00000014, pqac-00000017) | Valverde 2019 | https://doi.org/10.1083/jcb.201811139 | Rescue tested with ATG2A, but phenotype is defined in ATG2A/ATG2B DKO cells, supporting redundancy. |
| Interaction | A conserved ATG2 LIR mediates preferential binding to GABARAP-family proteins and is critical for phagophore maturation/closure. | Primary | Mammalian cells; biochemical binding assays | Mutation of the ATG2 LIR abolished ATG8-family binding and blocked phagophore formation/closure, phenocopying ATG2A/ATG2B DKO; starvation increased ATG2–GABARAP co-precipitation. (pqac-00000010, pqac-00000012) | Bozic 2020 | https://doi.org/10.15252/embr.201948412 | Demonstrated for both ATG2A and ATG2B family architecture; endogenous ATG2B co-localized with GFP-ATG2A. |
| Interaction | WIPI4/WDR45 binds mammalian ATG2 proteins, but ATG2–WIPI4 binding appears less essential than ATG2–GABARAP binding for autophagy flux in mammalian cells. | Primary + review | Mammalian cells; literature synthesis | WIPI4-binding mutants retained autophagy rescue, whereas GABARAP-binding mutants did not. (pqac-00000012, pqac-00000024) | Bozic 2020 | https://doi.org/10.15252/embr.201948412 | WIPI4 interacts with both ATG2A and ATG2B; functional importance may differ from yeast precedent. |
| Localization | ATG2 localizes to ER–phagophore/omegasome contact sites; N-terminus associates with ER and C-terminal regions contribute to phagophore binding. | Review | Mammalian/yeast literature synthesis | Overexpressed GFP-ATG2A localizes to phagophore rim and ER MCSs; coincidence binding to ATG9 and PI3P helps phagophore-edge association. (pqac-00000009) | Duarte 2023 | https://doi.org/10.1177/25152564231183898 | Localization model is derived mainly from ATG2A and yeast Atg2, but considered applicable to mammalian ATG2B as a redundant paralog. |
| Localization | Endogenous ATG2B and GFP-ATG2A co-localize on punctate/ring-like LC3B-positive starvation-induced structures together with early autophagy markers WIPI2 and ATG16L1. | Primary | Starved mammalian cells | Co-localization with LC3B, WIPI2, ATG16L1 on early autophagic structures. (pqac-00000010) | Bozic 2020 | https://doi.org/10.15252/embr.201948412 | One of the clearer pieces of direct ATG2B localization evidence in human/mammalian cells. |
| Phenotype | Simultaneous loss of ATG2A and ATG2B causes accumulation of small, open, immature phagophores and blocks autophagic flux. | Primary | Mammalian double depletion/KO systems | LC3B-II, GABARAPL1, and p62 accumulate; bafilomycin A1 fails to further increase LC3B-II in ATG2 DKO; large LC3-positive structures accumulate. (pqac-00000014, pqac-00000017, pqac-00000010) | Valverde 2019 | https://doi.org/10.1083/jcb.201811139 | Strongest functional evidence is for combined ATG2A/ATG2B loss, indicating partial redundancy. |
| Phenotype | ATG2A/B deficiency can shift stress responses away from cytoprotective autophagy toward apoptosis/caspase-8 activation. | Primary | Mammalian cells | Combined ATG2A/B loss caused immature autophagosomal membrane accumulation and altered death signaling. (context from paper search result; summarized in retrieved metadata) | Tang 2017 | https://doi.org/10.1038/cdd.2017.133 | Study addresses combined deficiency, not ATG2B alone. |
| New mechanism (2024) | ANKFY1 is a new ATG2A-binding factor that recruits/promotes ATG2-mediated lipid transfer from PI3P-positive endosomes to phagophores, identifying endosomes as an additional lipid source. | Primary | Mammalian cells; purified proteins/liposomes | ANKFY1 depletion impaired autophagosome growth and autophagy flux and largely phenocopied ATG2A/B depletion; recombinant ANKFY1 enhanced ATG2A-mediated lipid transfer; UVRAG, ANKFY1, or ATG2A/B depletion reduced PI3P on phagophores. (pqac-00000021) | Wei 2024 | https://doi.org/10.1038/s41421-024-00659-y | Direct binding was shown for ATG2A, but authors state ATG2A and ATG2B are functionally redundant in mammalian autophagy. |
| New mechanism (2023) | Lysosome damage triggers non-canonical ATG8 conjugation that engages ATG2 in a repair-associated pathway, extending ATG2 function beyond canonical autophagosome biogenesis. | Primary | Mammalian lysosome damage model | CASM generated direct ATG8 conjugation to lysosomal membranes and promoted robust LC3A engagement with ATG2. (pqac-00000027) | Cross 2023 | https://doi.org/10.1083/jcb.202303078 | Study discusses mammalian ATG2A/ATG2B collectively; isoform-specific distinctions were not emphasized. |
| Pathway context | ATG2 functions with ATG9A scramblase and WIPI/PI3P machinery at phagophore–ER contact sites; GABARAP interaction is especially important for late phagophore maturation. | Review | Autophagy pathway synthesis | Integrates lipid transfer, ATG9A scrambling, PI3P/WIPI scaffolding, and ATG8-family interactions in autophagosome biogenesis. (pqac-00000023, pqac-00000024) | Duarte 2023 | https://doi.org/10.1177/25152564231183898 | Pathway assignment applies to mammalian ATG2A and ATG2B together. |
| Disease genetics | A germline ~700-kb duplication at 14q32 including ATG2B and GSKIP segregates with familial myeloid malignancies. | Primary | Human pedigrees with familial MPN/AML spectrum | Duplication predisposed to ET/MPN and progression to AML/myelofibrosis/CMML; overexpression of ATG2B/GSKIP enhanced hematopoietic progenitor differentiation and cooperated with JAK2/MPL/CALR driver mutations. (pqac-00000028) | Saliba 2015 | https://doi.org/10.1038/ng.3380 | Disease association is not ATG2B-only because the duplicated interval also contains GSKIP. |
| Disease mechanism | In patient-derived models, increased ATG2B/GSKIP dosage promoted hematopoietic output and megakaryopoiesis. | Primary | Patient cells, iPSC differentiation, megakaryocyte assays | ~10-fold increase in colony formation in iPSC differentiation assays; two- to threefold overexpression in megakaryocytes; shRNA silencing reduced TPO-independent megakaryocyte formation and CFU-MK output to ~40% of maximal stimulated colonies. (pqac-00000029, pqac-00000030) | Saliba 2015 | https://doi.org/10.1038/ng.3380 | Mechanistic attribution remains shared between ATG2B and GSKIP. |
| Disease relevance/update | Mouse combined loss of Atg2b and Gskip impaired hematopoietic stem cell maintenance, supporting dosage-sensitive biology of the 14q32 locus, though effects were reported as autophagy-independent in that context. | Primary | Mouse knockout | Double knockout caused in utero lethality, decreased HSC pool size, and increased apoptosis; single knockouts showed little hematopoietic phenotype. (pqac-00000033, pqac-00000034) | Sakai 2022 | https://doi.org/10.1128/mcb.00024-21 | Again a locus-level effect (Atg2b + Gskip), not cleanly ATG2B-specific. |


*Table: This table summarizes the main functional, localization, interaction, phenotype, and disease-association evidence relevant to human ATG2B/Q96BY7. It emphasizes recent 2023-2024 studies while retaining foundational mechanistic and clinical genetics papers needed for a robust annotation.*