ACBD3 (GCP60, PAP7) is a peripheral membrane scaffolding protein of the Golgi and trans-Golgi network. It reaches Golgi membranes by a two-step route: an SCFD1-dependent step is required upstream, with ACBD3 binding the SNARE SEC22B at that step, and the golgins giantin and golgin-45 then bind redundantly to an MWT374-376 motif in the unique region immediately upstream of the GOLD domain. From there it organises lipid-modifying and signalling machinery on those membranes and at ER-Golgi contact sites. Its central activity is recruitment of the lipid kinase PI4KB through its Q domain (residues 241-308): the interaction anchors PI4KB to the membrane and stimulates the enzyme by raising the local concentration of its substrate, maintaining Golgi PI4P homeostasis. The same Q domain also binds the Rab GTPase-activating proteins TBC1D22A and TBC1D22B, and those interactions are mutually exclusive with PI4KB binding, making the domain a switch between lipid-kinase recruitment and Rab-directed traffic. The GOLD domain together with that same unique region forms the second interaction surface, and it is a shared one rather than a set of private sites: the SEC22B longin domain and the PKA type II regulatory subunit bind there, and so do picornaviral 3A proteins, whose contact residues lie in the unique region only a few residues from the golgin-binding motif. Because that surface is distinct from the Q domain, viral engagement leaves PI4KB in place; because it overlaps the golgin and PKA sites, the likely mechanism of hijack is competition with ACBD3's own Golgi anchors. ACBD3 acts as an A-kinase anchoring protein, binding PKA regulatory subunits to position the kinase at Golgi membranes, where it controls cargo-triggered PKA activation governing KDEL-receptor retrograde traffic. A second, mitochondrial pool has been proposed to act through the peripheral benzodiazepine receptor and to couple cholesterol transport to hormone-stimulated steroidogenesis, but it is inferred by similarity rather than localised directly in human cells. ACBD3 is required for Golgi stack integrity - its loss gives enlarged, unstacked Golgi - and for FAPP2-mediated glucosylceramide transport and ER-to-Golgi sphingolipid flux. The same membrane-coupling activity is extensively exploited by pathogens: picornavirus 3A proteins clamp ACBD3 onto replication-organelle membranes to recruit and activate PI4KB for viral phosphoinositide synthesis. The N-terminal acyl-CoA binding (ACB) domain is a third functional region: it binds long-chain fatty acyl-CoAs - C18:1-CoA and C16:0-CoA drive ACBD3 to oligomerise - and it has been assigned recruitment of the glycolipid-transfer adaptor FAPP2. Separately, ACBD3 restrains SREBP1 maturation. It binds the SREBP1 precursor but not the processed nuclear form, and depleting ACBD3 raises nuclear SREBP1 while ectopic ACBD3 suppresses it, lowering fatty-acid synthase expression and, on prolonged expression, de novo palmitate synthesis. Removing the N-terminus weakens that suppression without abolishing it, so the ACB region contributes to the effect rather than carrying it, and the SREBP1 interaction itself has only been assayed with full-length ACBD3. What the acyl-CoA ligand does remains open: deleting the ACB domain does not prevent Golgi localisation, so ligand-induced oligomerisation is not required for ACBD3 to reach the Golgi. The domain is also dispensable for 3A-mediated PI4KB recruitment, though that was assayed in enterovirus-infected cells rather than for host recruitment on its own.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0006888 endoplasmic reticulum to Golgi vesicle-mediated transport | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic inference that ACBD3 acts in ER-to-Golgi vesicle transport. Well founded and independently supported on this gene by an IMP: ACBD3 maintains Golgi structure through giantin, which affects protein transport between the ER and Golgi, and it governs cargo-triggered PKA activation controlling KDEL-receptor retrograde trafficking. Core process. Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt CC -!- FUNCTION: Involved in the maintenance of Golgi structure by interacting |
| GO:0007030 Golgi organization | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic inference of a role in Golgi organization, independently supported on this gene by an IDA. ACBD3 is required for Golgi stack integrity; knockout produces enlarged, unstacked Golgi. Core process. Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt CC -!- FUNCTION: Involved in the maintenance of Golgi structure by interacting |
| GO:0043495 protein-membrane adaptor activity | IBA GO_REF:0000033 | ACCEPT | Summary: Protein-membrane adaptor activity, by phylogenetic inference. This is the best molecular function on the gene and it is exactly right: ACBD3 is a peripheral membrane scaffold whose function is to bring soluble proteins to the Golgi membrane - PI4KB, PKA regulatory subunits, and, when hijacked, picornaviral 3A proteins. The IBA captures what six IPI rows record only as bare protein binding. Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt Recruits PI4KB to the Golgi apparatus membrane; enhances the enzyme file:human/ACBD3/ACBD3-deep-research-affinage.md Its central scaffolding output is the direct recruitment of the lipid kinase PI4KB to membranes |
| GO:0000139 Golgi membrane | IBA GO_REF:0000033 | ACCEPT | Summary: Golgi membrane. This is the protein's defining compartment: ACBD3 is a peripheral Golgi-membrane scaffold, targeted there in two steps - an SCFD1-dependent step upstream, then redundant binding of the golgins giantin and golgin-45 to an MWT374-376 motif in the unique region just upstream of the GOLD domain (PMID:38134218) - and it is where it recruits PI4KB and anchors PKA. Core location. Phylogenetic inference, and one of eight annotations to this term across five evidence codes - unusually well corroborated. Supporting Evidence: file:human/ACBD3/ACBD3-deep-research-affinage.md ACBD3 (GCP60/PAP7) is a peripheral Golgi-membrane scaffolding protein that organizes lipid-modifying and signaling machinery on Golgi/TGN membranes and at ERβGolgi contact sites PMID:38134218 We therefore concluded that the second mechanism for Golgi recruitment of ACBD3 is between the MWT374-376 residues of ACBD3 and two golgins: golgin-45 and giantin. PMID:38134218 This suggests that the recruitment of ACBD3 via SCFD1 is upstream of the interaction with the golgins, and ACBD3 is recruited to the Golgi in a two-step process. |
| GO:0000062 fatty-acyl-CoA binding | IEA GO_REF:0000002 | KEEP AS NON CORE | Summary: Fatty-acyl-CoA binding, assigned by InterPro from the ACB (acyl-CoA binding) domain at residues 83-174. Corrected call. Earlier versions of this review marked this over-annotated, on the grounds that no acyl-CoA ligand had ever been reported for ACBD3. That was wrong, and the correcting source was already cited here: PMID:38134218's introduction states that the ACBD domain oligomerizes upon binding C18:1-CoA or C16:0-CoA, citing Soupene and Kuypers 2015 (PMID:26290611), whose abstract records that ligand binding dimerises ACBD1 and ACBD3 while it does not dimerise ACBD6. So purified ACBD3 binds long-chain fatty acyl-CoAs, with ligand-induced oligomerisation as the readout - which is what the term asserts. The ACB region is also not functionless for this gene. PMID:23166793 shows ACBD3 blocks SREBP1 maturation by binding the regulator directly, with reduced FASN and reduced de novo palmitate synthesis following, and attributes an important part of that effect to the ACB-domain-containing N-terminal region. Only a part: deleting the N-terminus attenuates suppression of the FASN promoter (76% to 40%) without abolishing it, and the SREBP1 co-immunoprecipitation itself used full-length ACBD3, so the interaction is not domain-mapped. The ACB region therefore contributes to a gene-specific function that is distinct from acyl-CoA binding itself, without being where that function resides. KEEP_AS_NON_CORE rather than ACCEPT, and the reason is now the authors' own rather than my inference. PMID:38134218's discussion states that binding of ACBD3 to C18:1-CoA, C16:0-CoA or related acyl chains and the subsequent oligomerisation does not affect its recruitment to the Golgi, because deleting the ACBD domain does not prevent Golgi localisation - and lists effector recruitment among the possibilities that remain untested. Consistently, PMID:30755512 shows the ACB and CAR domains are dispensable for 3A-mediated PI4KB recruitment (assayed in the enteroviral setting, but scoring the Q-domain ACBD3-PI4KB interaction that the same paper shows is the step that matters), with Q and GOLD together sufficient. Note the framing: that assay is the viral one and host PI4KB recruitment without 3A was never tested, so it is corroboration rather than a second independent leg. Note this is a statement about the ligand, not about the domain. The ACB domain is not functionless: besides SREBP1 it is credited with recruiting FAPP2 (PMID:38134218 attributing PMID:29750412), so it is a third interaction surface alongside the Q domain and the UR/GOLD surface. What is absent is any demonstrated consequence of the acyl-CoA ligand itself. A real molecular function of this protein, but not its core one. Reason: Corrected from MARK_AS_OVER_ANNOTATED. Direct fatty-acyl-CoA binding by ACBD3 is reported - ligand-induced dimerisation with C18:1-CoA and C16:0-CoA (PMID:26290611, restated in PMID:38134218) - so the InterPro domain inference is corroborated, not speculative, and the ACB region additionally contributes to a gene-specific SREBP1-regulatory function - only in part, since the N-terminal deletion attenuates without abolishing it and the SREBP1 interaction was assayed only with full-length protein (PMID:23166793). Kept non-core on the authors' own evidence: PMID:38134218 states that acyl-CoA binding and the resulting oligomerisation do not affect ACBD3's Golgi recruitment, since deleting the ACBD domain does not prevent Golgi localisation, and PMID:30755512 shows the domain is dispensable for 3A-mediated PI4KB recruitment (assayed in the enteroviral setting, but scoring the Q-domain ACBD3-PI4KB interaction that the same paper shows is the step that matters). That bears on the ligand, not on the domain, which also recruits FAPP2. Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt FT DOMAIN 83..174 PMID:38134218 The ACBD domain oligomerizes upon binding to C18:1-CoA or C16:0-CoA (Soupene and Kuypers, 2015) and recruits the membrane-shaping protein FAPP2 PMID:38134218 ACBD3 has been demonstrated to oligomerize in the presence of C18:1-CoA or C16:0-CoA PMID:38134218 Our data suggest that the binding of ACBD3 to C18:1-CoA, C16:0-CoA or related fatty acyl chains and the subsequent oligomerization does not affect its recruitment to the Golgi as deletion of the ACBD domain does not prevent Golgi localization. PMID:38134218 It is possible that this oligomerization regulates the conformational change of ACBD3, its effector recruitment, or represents an additional functional role of ACBD3, which could be explored in future studies. PMID:26290611 In contrast to ACBD1 and ACBD3, ligand binding did not result in the dimerization of ACBD6. PMID:23166793 Taken together, these results suggest that ACB domain-containing N-terminal sequence of ACBD3 plays an important role in its regulatory effects on SREBP1. PMID:30755512 we show that acyl-coenzyme A binding (ACB) and charged-amino-acid region (CAR) domains are dispensable for 3A-mediated PI4KB recruitment and efficient enterovirus replication PMID:30755512 we dissected the different domains of ACBD3 and uncovered that the glutamine-rich region (Q) and Golgi dynamics domain (GOLD) together suffice to support enterovirus replication |
| GO:0000139 Golgi membrane | IEA GO_REF:0000120 | ACCEPT | Summary: Golgi membrane. This is the protein's defining compartment: ACBD3 is a peripheral Golgi-membrane scaffold, targeted there in two steps - an SCFD1-dependent step upstream, then redundant binding of the golgins giantin and golgin-45 to an MWT374-376 motif in the unique region just upstream of the GOLD domain (PMID:38134218) - and it is where it recruits PI4KB and anchors PKA. Core location. Automatic annotation combining multiple IEA methods. Supporting Evidence: file:human/ACBD3/ACBD3-deep-research-affinage.md ACBD3 (GCP60/PAP7) is a peripheral Golgi-membrane scaffolding protein that organizes lipid-modifying and signaling machinery on Golgi/TGN membranes and at ERβGolgi contact sites PMID:38134218 We therefore concluded that the second mechanism for Golgi recruitment of ACBD3 is between the MWT374-376 residues of ACBD3 and two golgins: golgin-45 and giantin. PMID:38134218 This suggests that the recruitment of ACBD3 via SCFD1 is upstream of the interaction with the golgins, and ACBD3 is recruited to the Golgi in a two-step process. |
| GO:0005739 mitochondrion | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: Mitochondrion, from the UniProt SubCell mapping. Genuine but secondary, and inferred rather than observed in human: UniProt records the mitochondrial pool as "Mitochondrion {ECO:0000250}. Note=Also mitochondrial (via its interaction with PBR)", i.e. by similarity with no human experimental localisation. Functionally it is where ACBD3 positions PKA to couple cholesterol transport to hormone-stimulated steroidogenesis in Leydig cells, and the type I PKA regulatory subunit is itself the mitochondrially enriched one. The Golgi pool is the protein's principal location, the site of its defining scaffolding activity, and the only one with human experimental localisation, so this is retained as non-core. Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt Mitochondrion {ECO:0000250}. Note=Also mitochondrial PMID:37044218 The discrepancy between ACBD3βs preference for the R subunit at different organelles may be explained by the fact that RI and RII are highly enriched at the mitochondria and the Golgi apparatus, respectively |
| GO:0005515 protein binding | IPI PMID:22124328 ACBD3-mediated recruitment of PI4KB to picornavirus RNA repl... | MODIFY | Summary: Interaction with PI4KB (UniProtKB:Q9UBF8) and with the Aichi virus 3A protein and its processed chains (UniProtKB:O91464, four of this row's five WITH/FROM accessions) - the paper recruits both to the same replication-site question. PI4KB is the central host partner of this protein and the one its molecular function is built on. ACBD3 binds PI4KB through its Q domain (residues 241-308), anchoring the kinase to Golgi membranes and stimulating its activity by raising the local substrate concentration, thereby maintaining Golgi PI4P homeostasis; the complex was defined structurally by NMR of ACBD3 residues 241-308. Bare protein binding conveys none of this. Note the domain assignment specifically, since it is easy to get wrong: ACBD3 is named for its acyl-CoA-binding domain, and the GOLD domain is the one usually discussed, but neither carries the PI4KB interaction. UniProt is explicit that the Q domain binds PI4KB and TBC1D22A/B, while the GOLD domain (384-526) binds picornaviral 3A proteins. The viral 3A interaction covered by this same row therefore uses a different surface from the host kinase interaction - which is why 3A can clamp ACBD3 onto replication organelles without displacing PI4KB. Reason: Uninformative term for the structurally defined interaction that constitutes this protein's molecular function. Proposed replacements: protein-membrane adaptor activity Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt CC similarity). Interacts (via Q domain) with PI4KB (via N-terminus) file:human/ACBD3/ACBD3-uniprot.txt CC -!- SUBUNIT: (Microbial infection) Interacts (via GOLD domain) with 3A file:human/ACBD3/ACBD3-uniprot.txt Recruits PI4KB to the Golgi apparatus membrane; enhances the enzyme |
| GO:0005515 protein binding | IPI PMID:22796112 Acyl-CoA binding domain containing 3 (ACBD3) recruits the pr... | MARK AS OVER ANNOTATED | Summary: Interaction with PPM1L (mouse Ppm1l, UniProtKB:Q8BHN0), a phosphatase acting in ER-to-Golgi ceramide transport. Topically coherent - ACBD3 is required for FAPP2-mediated glucosylceramide transport and ER-to-Golgi sphingolipid flux - but this specific pair has no dedicated follow-up, and the term is uninformative. Marked over-annotated rather than elevated, since the sphingolipid role is established independently of this interaction. Reason: Uninformative term for an interaction with no dedicated follow-up; the related sphingolipid function is established by other evidence. Supporting Evidence: file:human/ACBD3/ACBD3-deep-research-affinage.md ACBD3 (GCP60/PAP7) is a peripheral Golgi-membrane scaffolding protein that organizes lipid-modifying and signaling machinery on Golgi/TGN membranes and at ERβGolgi contact sites |
| GO:0005515 protein binding | IPI PMID:23572552 ACBD3 interaction with TBC1 domain 22 protein is differentia... | MODIFY | Summary: This row collapses nine IntAct/UniProt IPI rows from a single paper, and the WITH/FROM accessions are what make it interpretable: PI4KB (Q9UBF8, plus the unreviewed isoform A0A0B4J1S8), TBC1D22A and TBC1D22B (Q8WUA7, Q9NU19), and the 3A regions of three picornaviral polyproteins - Aichi virus (O91464, a kobuvirus), poliovirus type 1 Mahoney (P03300) and coxsackievirus B3 Woodruff (Q66282), both enteroviruses. That set is the paper's argument, not a screen dragnet: it is titled for the observation that ACBD3-TBC1D22 binding is affected differently by enteroviral than by kobuviral 3A. Two mechanisms sit behind it, on two different surfaces. PI4KB and TBC1D22A/B compete for the same Q domain (residues 241-308) - UniProt states the two interactions are mutually exclusive - so that domain is a switch between lipid-kinase recruitment and Rab-directed traffic. The viral 3A proteins bind elsewhere, at the GOLD domain and the unique region upstream of it, which is why 3A can clamp ACBD3 without displacing PI4KB while still modulating the Rab-GAP arm. Bare protein binding records none of this, and it is the same MODIFY as the other uninformative rows: every partner here engages ACBD3 as a membrane-recruitment adaptor. Reason: Uninformative term for a paper whose whole point is which surface each partner uses: PI4KB and TBC1D22A/B competing for the Q domain, and enteroviral versus kobuviral 3A proteins engaging the GOLD/unique-region surface with different consequences for the Rab-GAP arm. All nine collapsed WITH/FROM accessions were resolved and all are real, mechanistically relevant partners, so this is a term problem rather than an evidence problem. Proposed replacements: protein-membrane adaptor activity Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt CC similarity). Interacts (via Q domain) with PI4KB (via N-terminus) file:human/ACBD3/ACBD3-uniprot.txt PubMed:27989622). Interacts (via Q domain) with TBC1D22A and TBC1D22B; file:human/ACBD3/ACBD3-uniprot.txt CC -!- SUBUNIT: (Microbial infection) Interacts (via GOLD domain) with 3A PMID:38134218 Picornavirus 3A peptide recruits ACBD3 to viral replication sites through a proteinβprotein interaction via the UR of the GOLD domain of ACBD3 |
| GO:0005515 protein binding | IPI PMID:28514442 Architecture of the human interactome defines protein commun... | MODIFY | Summary: Interaction with TBC1D22B (UniProtKB:Q9NU19 - this row's only WITH/FROM accession; the companion PMID:33961781 row carries both TBC1D22A and TBC1D22B), a Rab GTPase-activating protein. These are easy to mistake for uncharacterised screen hits, but UniProt shows the interaction is mechanistically pointed. TBC1D22A/B bind the SAME Q domain as PI4KB, and UniProt states the two interactions are mutually exclusive. So ACBD3's Q domain is a switch: it can hold either the lipid kinase or the Rab GAPs, not both. That makes these rows worth elevating rather than flagging. They record engagement of the same adaptor surface that defines the protein's molecular function, and the competition is presumably how the Golgi decides between PI4P synthesis and Rab-directed traffic. Bare protein binding records none of it. Reason: Uninformative term for an interaction at the same Q-domain surface that mediates PI4KB recruitment, and mutually exclusive with it - a documented competitive switch rather than an unexplained hit. Proposed replacements: protein-membrane adaptor activity Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt PubMed:27989622). Interacts (via Q domain) with TBC1D22A and TBC1D22B; file:human/ACBD3/ACBD3-uniprot.txt CC similarity). Interacts (via Q domain) with PI4KB (via N-terminus) |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | MODIFY | Summary: Interaction with TBC1D22A/TBC1D22B (UniProtKB:Q8WUA7, Q9NU19), Rab GTPase-activating proteins. These are easy to mistake for uncharacterised screen hits, but UniProt shows the interaction is mechanistically pointed. TBC1D22A/B bind the SAME Q domain as PI4KB, and UniProt states the two interactions are mutually exclusive. So ACBD3's Q domain is a switch: it can hold either the lipid kinase or the Rab GAPs, not both. That makes these rows worth elevating rather than flagging. They record engagement of the same adaptor surface that defines the protein's molecular function, and the competition is presumably how the Golgi decides between PI4P synthesis and Rab-directed traffic. Bare protein binding records none of it. Reason: Uninformative term for an interaction at the same Q-domain surface that mediates PI4KB recruitment, and mutually exclusive with it - a documented competitive switch rather than an unexplained hit. Proposed replacements: protein-membrane adaptor activity Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt PubMed:27989622). Interacts (via Q domain) with TBC1D22A and TBC1D22B; file:human/ACBD3/ACBD3-uniprot.txt CC similarity). Interacts (via Q domain) with PI4KB (via N-terminus) |
| GO:0006888 endoplasmic reticulum to Golgi vesicle-mediated transport | IMP PMID:37044218 An A-kinase anchoring protein (ACBD3) coordinates traffic-in... | ACCEPT | Summary: ER-to-Golgi vesicle-mediated transport by mutant phenotype, from PMID:37044218 - the experimental counterpart of the IBA above. Worth being precise about what the paper shows, because the term names the anterograde leg while the paper's subject is the retrograde one. ACBD3 holds the PKA holoenzyme at the Golgi through RII. Arrival of anterograde cargo from the ER releases the catalytic subunit, activating PKA; ACBD3 depletion reduces Golgi RII and makes that activation - and KDEL-receptor retrograde transport - constitutive and independent of cargo influx. So ACBD3 is the element that couples the two directions, making retrograde recycling responsive to forward flux. ACCEPT: the loss-of-function phenotype does perturb traffic between ER and Golgi, and ACBD3 acts in the early secretory pathway, so the term is not wrong. But it captures the trigger rather than the regulated process, which is why a retrograde regulation term is proposed as a NEW annotation below - additively, not as a replacement for a curator's experimental call. Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt CC -!- FUNCTION: Involved in the maintenance of Golgi structure by interacting PMID:37044218 Forward trafficking of proteins from the ER triggers activation of PKA by releasing the catalytic subunit from RII PMID:37044218 depletion of ACBD3 reduces the Golgi fraction of RII, resulting in moderate, but constitutive activation of PKA and KDELR retrograde transport, independent of cargo influx from the ER |
| GO:0007030 Golgi organization | IDA PMID:11590181 Identification and characterization of a novel Golgi protein... | ACCEPT | Summary: Golgi organization by direct assay, from the paper that identified ACBD3 as a giantin-binding protein. ACBD3 maintains Golgi structure through that interaction, and knockout produces enlarged, unstacked Golgi. Core process, and the oldest well-supported annotation on the gene. Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt CC -!- FUNCTION: Involved in the maintenance of Golgi structure by interacting file:human/ACBD3/ACBD3-deep-research-affinage.md ACBD3 is itself targeted to the Golgi by a two-step mechanism |
| GO:0005794 Golgi apparatus | IDA GO_REF:0000052 | KEEP AS NON CORE | Summary: Golgi apparatus by direct immunofluorescence (Human Protein Atlas). Correct, and the parent of the Golgi membrane term annotated eight times over; retained as the less specific of the two. Supporting Evidence: file:human/ACBD3/ACBD3-deep-research-affinage.md ACBD3 (GCP60/PAP7) is a peripheral Golgi-membrane scaffolding protein that organizes lipid-modifying and signaling machinery on Golgi/TGN membranes and at ERβGolgi contact sites |
| GO:0000139 Golgi membrane | EXP PMID:22124328 ACBD3-mediated recruitment of PI4KB to picornavirus RNA repl... | ACCEPT | Summary: Golgi membrane. This is the protein's defining compartment: ACBD3 is a peripheral Golgi-membrane scaffold, targeted there in two steps - an SCFD1-dependent step upstream, then redundant binding of the golgins giantin and golgin-45 to an MWT374-376 motif in the unique region just upstream of the GOLD domain (PMID:38134218) - and it is where it recruits PI4KB and anchors PKA. Core location. Direct experimental localisation. Supporting Evidence: file:human/ACBD3/ACBD3-deep-research-affinage.md ACBD3 (GCP60/PAP7) is a peripheral Golgi-membrane scaffolding protein that organizes lipid-modifying and signaling machinery on Golgi/TGN membranes and at ERβGolgi contact sites PMID:38134218 We therefore concluded that the second mechanism for Golgi recruitment of ACBD3 is between the MWT374-376 residues of ACBD3 and two golgins: golgin-45 and giantin. PMID:38134218 This suggests that the recruitment of ACBD3 via SCFD1 is upstream of the interaction with the golgins, and ACBD3 is recruited to the Golgi in a two-step process. |
| GO:0000139 Golgi membrane | EXP PMID:30755512 ACBD3 Is an Essential Pan-enterovirus Host Factor That Media... | ACCEPT | Summary: Golgi membrane. This is the protein's defining compartment: ACBD3 is a peripheral Golgi-membrane scaffold, targeted there in two steps - an SCFD1-dependent step upstream, then redundant binding of the golgins giantin and golgin-45 to an MWT374-376 motif in the unique region just upstream of the GOLD domain (PMID:38134218) - and it is where it recruits PI4KB and anchors PKA. Core location. Direct experimental localisation. Supporting Evidence: file:human/ACBD3/ACBD3-deep-research-affinage.md ACBD3 (GCP60/PAP7) is a peripheral Golgi-membrane scaffolding protein that organizes lipid-modifying and signaling machinery on Golgi/TGN membranes and at ERβGolgi contact sites PMID:38134218 We therefore concluded that the second mechanism for Golgi recruitment of ACBD3 is between the MWT374-376 residues of ACBD3 and two golgins: golgin-45 and giantin. PMID:38134218 This suggests that the recruitment of ACBD3 via SCFD1 is upstream of the interaction with the golgins, and ACBD3 is recruited to the Golgi in a two-step process. |
| GO:0000139 Golgi membrane | EXP PMID:31381608 Convergent evolution in the mechanisms of ACBD3 recruitment ... | ACCEPT | Summary: Golgi membrane. This is the protein's defining compartment: ACBD3 is a peripheral Golgi-membrane scaffold, targeted there in two steps - an SCFD1-dependent step upstream, then redundant binding of the golgins giantin and golgin-45 to an MWT374-376 motif in the unique region just upstream of the GOLD domain (PMID:38134218) - and it is where it recruits PI4KB and anchors PKA. Core location. Direct experimental localisation. Supporting Evidence: file:human/ACBD3/ACBD3-deep-research-affinage.md ACBD3 (GCP60/PAP7) is a peripheral Golgi-membrane scaffolding protein that organizes lipid-modifying and signaling machinery on Golgi/TGN membranes and at ERβGolgi contact sites PMID:38134218 We therefore concluded that the second mechanism for Golgi recruitment of ACBD3 is between the MWT374-376 residues of ACBD3 and two golgins: golgin-45 and giantin. PMID:38134218 This suggests that the recruitment of ACBD3 via SCFD1 is upstream of the interaction with the golgins, and ACBD3 is recruited to the Golgi in a two-step process. |
| GO:0000139 Golgi membrane | IDA PMID:27009356 Structural insights and in vitro reconstitution of membrane ... | ACCEPT | Summary: Golgi membrane. This is the protein's defining compartment: ACBD3 is a peripheral Golgi-membrane scaffold, targeted there in two steps - an SCFD1-dependent step upstream, then redundant binding of the golgins giantin and golgin-45 to an MWT374-376 motif in the unique region just upstream of the GOLD domain (PMID:38134218) - and it is where it recruits PI4KB and anchors PKA. Core location. Direct experimental localisation. Supporting Evidence: file:human/ACBD3/ACBD3-deep-research-affinage.md ACBD3 (GCP60/PAP7) is a peripheral Golgi-membrane scaffolding protein that organizes lipid-modifying and signaling machinery on Golgi/TGN membranes and at ERβGolgi contact sites PMID:38134218 We therefore concluded that the second mechanism for Golgi recruitment of ACBD3 is between the MWT374-376 residues of ACBD3 and two golgins: golgin-45 and giantin. PMID:38134218 This suggests that the recruitment of ACBD3 via SCFD1 is upstream of the interaction with the golgins, and ACBD3 is recruited to the Golgi in a two-step process. |
| GO:0005515 protein binding | IPI PMID:27009356 Structural insights and in vitro reconstitution of membrane ... | MODIFY | Summary: Interaction with PI4KB (UniProtKB:Q9UBF8), the central partner of this protein and the one its molecular function is built on. ACBD3 binds PI4KB through its Q domain (residues 241-308), anchoring the kinase to Golgi membranes and stimulating its activity by raising the local substrate concentration, thereby maintaining Golgi PI4P homeostasis; the complex was defined structurally by NMR of ACBD3 residues 241-308. Bare protein binding conveys none of this. Note the domain assignment specifically, since it is easy to get wrong: ACBD3 is named for its acyl-CoA-binding domain, and the GOLD domain is the one usually discussed, but neither carries the PI4KB interaction. UniProt is explicit that the Q domain binds PI4KB and TBC1D22A/B, while the GOLD domain (384-526) binds picornaviral 3A proteins. The viral 3A interaction covered by this same row therefore uses a different surface from the host kinase interaction - which is why 3A can clamp ACBD3 onto replication organelles without displacing PI4KB. Reason: Uninformative term for the structurally defined interaction that constitutes this protein's molecular function. Proposed replacements: protein-membrane adaptor activity Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt CC similarity). Interacts (via Q domain) with PI4KB (via N-terminus) file:human/ACBD3/ACBD3-uniprot.txt CC -!- SUBUNIT: (Microbial infection) Interacts (via GOLD domain) with 3A file:human/ACBD3/ACBD3-uniprot.txt Recruits PI4KB to the Golgi apparatus membrane; enhances the enzyme |
| GO:0000139 Golgi membrane | TAS Reactome:R-HSA-8874979 | ACCEPT | Summary: Golgi membrane. This is the protein's defining compartment: ACBD3 is a peripheral Golgi-membrane scaffold, targeted there in two steps - an SCFD1-dependent step upstream, then redundant binding of the golgins giantin and golgin-45 to an MWT374-376 motif in the unique region just upstream of the GOLD domain (PMID:38134218) - and it is where it recruits PI4KB and anchors PKA. Core location. From a Reactome reaction. Correct compartment, asserted at pathway level. Supporting Evidence: file:human/ACBD3/ACBD3-deep-research-affinage.md ACBD3 (GCP60/PAP7) is a peripheral Golgi-membrane scaffolding protein that organizes lipid-modifying and signaling machinery on Golgi/TGN membranes and at ERβGolgi contact sites PMID:38134218 We therefore concluded that the second mechanism for Golgi recruitment of ACBD3 is between the MWT374-376 residues of ACBD3 and two golgins: golgin-45 and giantin. PMID:38134218 This suggests that the recruitment of ACBD3 via SCFD1 is upstream of the interaction with the golgins, and ACBD3 is recruited to the Golgi in a two-step process. |
| GO:0034237 protein kinase A regulatory subunit binding | IPI PMID:17911601 Inhibition of T cell activation by cyclic adenosine 5'-monop... | ACCEPT | Summary: Protein kinase A regulatory subunit binding - an informative binding term, and the model for how the six bare protein-binding rows on this gene should look. The term is right and the anchoring function is well supported; the apparent argument over which R subunit is bound is largely reconciled inside the primary literature itself. The GOA row is an IPI whose WITH/FROM is UniProtKB:P10644 = PRKAR1A, so the curated interaction is with the type I-alpha regulatory subunit (RIalpha). UniProt's SUBUNIT line agrees and adds a negative: it states ACBD3 interacts with PBR and RI-alpha and does NOT interact with RI-beta nor RII-alpha - but that whole statement carries (By similarity), i.e. ECO:0000250, not experimental evidence for human ACBD3. PMID:37044218, a human experimental study, detects RIIalpha and not RIalpha in a GST-ACBD3 pull-down and maps the interaction to the GOLD domain. Its authors do not read that as excluding RI: they note that a typical dual-specific AKAP binds RI 10-100 fold more weakly than RII and that ACBD3 may bind RIalpha below their detection limit, and they attribute the organelle asymmetry to RI and RII being enriched at mitochondria and the Golgi respectively. On that reading the Golgi pool is RII-anchored and the mitochondrial pool RI-anchored, which is compatible with both records rather than in conflict with either. The same paper also sharpens the domain picture: the GOLD domain has host partners, RII here and the SEC22B longin domain in PMID:38134218, so it is not a viral-only surface. ACCEPT: GO:0034237 is agnostic about isoform, so it holds under any of these readings, and the curated IPI is an experimental annotation whose full text has not been read here. The one claim that does not survive the reconciliation is UniProt's (By similarity) negative for RII-alpha, which a human pull-down contradicts directly; that is flagged in suggested_questions as a UniProt correction to report. Reason: The term is correct under any reading of the subunit preference, and the curated IPI against UniProtKB:P10644 (PRKAR1A) is an experimental annotation whose full text I have not read, so it is not second-guessed. PMID:37044218 reconciles its own RII result with RI binding (affinity below detection; RI and RII enriched at mitochondria and Golgi respectively), so the record is reported as compartment-specific rather than contested. Only UniProt's by-similarity RII-alpha negative is flagged. Supporting Evidence: file:human/ACBD3/ACBD3-uniprot.txt PBR and PKA regulatory subunit RI-alpha PMID:37044218 His-RIIΞ±, but not His-RIΞ±, was readily detected in the pulldown fraction PMID:37044218 the GOLD domain of ACBD3 directly interacts with the regulatory subunit II (RII) of PKA and effectively recruits PKA holoenzyme to the Golgi PMID:37044218 it is possible that ACBD3 binds RIΞ± in such low affinity that this binding is below the detection limit in our assay. PMID:37044218 The discrepancy between ACBD3βs preference for the R subunit at different organelles may be explained by the fact that RI and RII are highly enriched at the mitochondria and the Golgi apparatus, respectively |
| GO:0016020 membrane | HDA PMID:19946888 Defining the membrane proteome of NK cells. | MARK AS OVER ANNOTATED | Summary: Generic membrane, from a high-throughput dataset. Uninformative when the specific Golgi membrane is annotated eight times over with EXP, IDA, IBA, IEA and TAS evidence. Reason: Redundant unspecified parent of the extensively supported Golgi membrane location. Supporting Evidence: file:human/ACBD3/ACBD3-deep-research-affinage.md ACBD3 (GCP60/PAP7) is a peripheral Golgi-membrane scaffolding protein that organizes lipid-modifying and signaling machinery on Golgi/TGN membranes and at ERβGolgi contact sites |
| GO:2000156 regulation of retrograde vesicle-mediated transport, Golgi to ER | IMP PMID:37044218 An A-kinase anchoring protein (ACBD3) coordinates traffic-in... | NEW | Summary: Proposed new annotation. ACBD3 has no retrograde-transport annotation at all, yet regulating retrograde traffic is what PMID:37044218 actually establishes - the GOA record captures only the anterograde term the same paper supports less directly. The evidence is a loss-of-function phenotype on the process itself: depleting ACBD3 reduces the Golgi pool of PKA RII and renders PKA activation and KDEL-receptor retrograde transport constitutive rather than cargo-triggered. The term's definition - any process that modulates the frequency, rate or extent of retrograde vesicle-mediated transport, Golgi to ER - fits that directly, and GO:2000156 has no children, so it is the most specific available term. The regulation parent is the right choice over GO:0006890 itself: ACBD3 is not a transport machinery component, it sets whether the pathway runs constitutively. Reason: Fills a real gap rather than restating an existing row: the paper's own conclusion is about retrograde transport control, and no annotation on this gene records it. IMP because the evidence is the ACBD3-depletion phenotype. Supporting Evidence: PMID:37044218 depletion of ACBD3 reduces the Golgi fraction of RII, resulting in moderate, but constitutive activation of PKA and KDELR retrograde transport, independent of cargo influx from the ER PMID:37044218 these data demonstrate that ACBD3 coordinates the protein secretory pathway at the Golgi by facilitating KDELR/PKA-containing protein complex formation |
| GO:0000149 SNARE binding | IPI PMID:38134218 Recruitment of PI4KIIIΞ² to the Golgi by ACBD3 is dependent o... | NEW | Summary: Proposed new annotation. ACBD3 binds the v-SNARE SEC22B, and the interaction is mapped on both sides: co-immunoprecipitation of GFP-ACBD3 with HaloTag-SEC22B, narrowed by truncation to ACBD3 328-528 (the unique region plus the GOLD domain) and, on SEC22B, to the longin domain alone - the SNARE and transmembrane domains do not bind. This is worth annotating for two reasons. First, the gene has no informative molecular-function term for any SNARE-pathway interaction, while six of its interactions sit under bare GO:0005515; SNARE binding is exactly the kind of specific term those rows should have. Second, it settles the question of whether the GOLD domain is a viral-only surface: it now has two host partners, the SEC22B longin domain here and PKA RIIalpha in PMID:37044218, alongside picornaviral 3A. Deliberately narrower than the paper's own wording. The authors conclude ACBD3 "is part of the SNARE complex involving the SM protein SCFD1 and the v-SNARE SEC22B", but the assays are binary co-IPs from overexpressing HEK-293T cells, and the cytosolic MWT374-376>AAA mutant binds SEC22B more strongly than wild type - so complex membership is not established and no complex or fusion term is proposed, only the binding. Two limits worth stating. The SEC22B knockout in the same screen "caused a drastic loss of Golgi organization resulting in Golgi fragmentation", so SEC22B's specific requirement for ACBD3 recruitment is confounded by a general Golgi defect - another reason to annotate the binding and not a requirement. And this is not proposed as a core function: the interaction serves ACBD3's own delivery to the cis-Golgi, upstream of what ACBD3 then does for the cell, whereas core_functions record its two outputs - PI4KB recruitment and PKA anchoring. An informative molecular-function term for a real, surface-mapped interaction does not have to be a core function. Reason: Fills a real gap: SEC22B is an ACBD3 partner with the interaction surface mapped on both proteins, and no annotation on this gene records any SNARE-pathway interaction. IPI because the evidence is reciprocal co-immunoprecipitation with domain mapping. Restricted to the binding term rather than complex membership, which the co-IP data do not establish. Supporting Evidence: PMID:38134218 We thus conclude that the UR and GOLD domain of ACBD3 interacts with the longin domain of SEC22B PMID:38134218 The SNARE and transmembrane domain did not interact with ACBD3 (328-528); however, the longin domain alone does interact with ACBD3 (328-528). PMID:38134218 Loss of TMED10 and SEC22B caused a drastic loss of Golgi organization resulting in Golgi fragmentation. |
| GO:0034067 protein localization to Golgi apparatus | IMP PMID:38134218 Recruitment of PI4KIIIΞ² to the Golgi by ACBD3 is dependent o... | NEW | Summary: Proposed new annotation. ACBD3's whole documented output is bringing other proteins to Golgi membranes - PI4KB, PKA regulatory subunits, FAPP2, PPM1L - yet the gene carries no protein-localisation process term at all. GO:0043495 protein-membrane adaptor activity records the molecular function; this is its biological-process counterpart. The evidence is a loss-of-function phenotype with the direction controlled: CRISPR-knockout of ACBD3 causes loss of PI4KIIIbeta from the Golgi apparatus, whereas knockout of PI4KIIIbeta leaves ACBD3 localisation unaffected. The same paper shows the dependency propagates upstream - SCFD1 knockout loses both ACBD3 and PI4KIIIbeta - which is what places ACBD3 between the two. GO:0034067 rather than a regulation term: ACBD3 is the adaptor that does the localising, not a modulator of someone else's localisation machinery. Reason: Fills a real gap: the gene's defining activity is recruiting client proteins to the Golgi, the molecular function is annotated but the process is not. IMP because the evidence is the ACBD3-knockout phenotype, with the reciprocal PI4KB knockout as the direction control. Supporting Evidence: PMID:38134218 the abrogation of ACBD3 caused a loss of localization of PI4KIIIΞ² to the Golgi apparatus PMID:38134218 CRISPR-KO of PI4KIIIΞ² did not affect the localization of ACBD3 file:human/ACBD3/ACBD3-uniprot.txt Recruits PI4KB to the Golgi apparatus membrane; enhances the enzyme |
| GO:2000639 negative regulation of SREBP signaling pathway | IMP PMID:23166793 Maturation and activity of sterol regulatory element binding... | NEW | Summary: Proposed new annotation, and the answer to a fair question: this review now asserts an ACB-domain SREBP1 function in the biologist-facing description, so it should either annotate it or say why not. The evidence is a coherent set from one paper. ACBD3 co-immunoprecipitates the full-length SREBP1 precursor but not the mature nuclear form; it blocks SREBP1 maturation by binding the regulator directly rather than by disrupting the SREBP1-SCAP-Insig1 complex; it attenuates the SREBP1 response to lipid deprivation; the ACB-domain-containing N-terminal region plays an important part in the effect, though deleting it attenuates rather than abolishes suppression (76% to 40%); and FASN transcript and protein fall, with de novo palmitate synthesis reduced at 72 h. The perturbation is bidirectional, which is the strongest form IMP takes. Overexpression suppresses SREBP1 maturation; shRNA knockdown of endogenous ACBD3 raises nuclear SREBP1 in two cell lines, HEK293T and HepG2, and the authors conclude ACBD3 "intrinsically plays a negative role in SREBP1 protein maturation". A specificity control comes with it - knocking down ACBD3 does not change the SREBP-binding partners SCAP or Insig1. IMP rather than IDA or IPI because both directions are perturbations rather than direct assays of the process. Still one paper, so proposed rather than asserted - but not thin: two directions, two cell lines, a lipid-deprivation challenge, a specificity control and a domain-deletion series. Not proposed as a core function, for the same reason as GO:0000149. core_functions records this protein's two scaffolding outputs - PI4KB recruitment at the Golgi and PKA anchoring - and SREBP1 restraint is neither: it is a separate, lipogenesis-facing arm running through the ACB region, on single-paper evidence. An informative annotation need not be a core function, and saying so is cheaper than leaving a reader to infer it. Reason: Fills a gap the narrative already opens: the review states an ACB-domain SREBP1-regulatory function, and no annotation on this gene records any role in lipogenic transcriptional control. GO:2000639 is the specific term for the direction shown - ACBD3 blocks SREBP1 maturation and attenuates the lipid-deprivation response. IMP because the evidence is perturbation rather than direct assay - and it is bidirectional: overexpression suppresses SREBP1 maturation, knockdown of endogenous ACBD3 enhances it in two cell lines, with a specificity control. Single-paper, and flagged as such. Supporting Evidence: PMID:23166793 ACBD3 blocked intracellular maturation of SREBP1 probably through directly binding with the lipid regulator rather than disrupted SREBP1-SCAP-Insig1 interaction PMID:23166793 When ACBD3-overexpressing cells were exposed to the lipid-deprivation condition, a prominent attenuation of the depletion-initiated SREBP1 feedback was observed PMID:23166793 Taken together, these results suggest that ACB domain-containing N-terminal sequence of ACBD3 plays an important role in its regulatory effects on SREBP1. PMID:23166793 More strikingly, after knocking down endogenous ACBD3, nuclear SREBP1 expression was enhanced in both HEK293T and Hep G2 cells PMID:23166793 All these results strongly suggest that ACBD3 intrinsically plays a negative role in SREBP1 protein maturation. PMID:23166793 knockdown of endogenous ACBD3 had no effect on the expression of these two SREBP-binding proteins, either (Figure S2) |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: What is the physiological consequence of acyl-CoA binding by ACBD3? The ACB domain binds C18:1-CoA and C16:0-CoA and oligomerises on ligand, and the ACB-containing N-terminal region restrains SREBP1 maturation - so the domain is neither vestigial nor silent. But no acyl-CoA-dependent step in ACBD3's Golgi scaffolding, PI4KB recruitment or AKAP function has been identified, and the ACB and CAR domains are dispensable for 3A-mediated PI4KB recruitment (the enteroviral assay; host recruitment without 3A was not tested) while Q and GOLD together suffice. Does ligand-induced oligomerisation gate any Golgi function - dimer valency would double the adaptor's client capacity - or is acyl-CoA sensing a separate, lipogenesis-facing arm of the protein?
Q: Why do six protein-binding rows record PI4KB, TBC1D22A/B and viral 3A proteins as bare GO:0005515 when the gene already carries an informative protein-membrane adaptor term and an informative PKA-regulatory-subunit-binding term? The uninformative rows are the ones from the focused mechanistic papers, which is the opposite of what one would expect.
Q: What does ACBD3 do with TBC1D22A and TBC1D22B? These Rab GAPs are recovered by interactome screens and appear in the picornavirus work as part of the ACBD3-organised complex, but no study has characterised the functional consequence of the interaction.
Q: Is the mitochondrial pool a distinct targeting event or spillover from the Golgi pool? ACBD3 positions PKA at both compartments, but both of its characterised targeting mechanisms - the SCFD1-dependent step and the MWT374-376 motif binding giantin and golgin-45 - are Golgi-specific, and UniProt's mitochondrial statement is (By similarity) via PBR with no human experimental localisation. PMID:37044218 suggests the answer may be trivial - RI and RII are themselves enriched at mitochondria and Golgi respectively - but no study has tested how ACBD3 itself reaches mitochondria.
Q: Should UniProt's (By similarity) negative for RII-alpha be revised, and is the R-subunit preference simply compartmental? The curated GO:0034237 IPI is against PRKAR1A (RIalpha, UniProtKB:P10644) and UniProt states RI-alpha binding with no RII-alpha binding - but only (By similarity). PMID:37044218 detects RIIalpha and not RIalpha in a human pull-down, yet its authors explicitly do not exclude RI: they note a dual-specific AKAP binds RI 10-100 fold more weakly and that RI and RII are enriched at mitochondria and the Golgi respectively, so the two records may simply describe one protein at two compartments. What does not survive that reconciliation is a by-similarity negative for RII-alpha standing against a human experimental positive, which looks like a UniProt correction to report.
Q: Is ACBD3 a genuine subunit of the STX5-SEC22B SNARE complex or a client recruited alongside it? PMID:38134218 shows ACBD3 binds the SEC22B longin domain through its unique region and GOLD domain and that SCFD1 loss strips ACBD3 from the Golgi, and concludes ACBD3 is part of that SNARE complex - but the MWT374-376>AAA mutant and the UR-less truncation bind SCFD1 and SEC22B more strongly than wild type while sitting in the cytosol, so binding and Golgi residence are separable. Whether ACBD3 influences SNARE-mediated fusion, or is only chaperoned to the cis-Golgi by it, is untested.
Experiment: Ligand binding and ligand-induced dimerisation are established for ACBD3; what is missing is any downstream consequence. Determine affinities for a short-, medium- and long-chain acyl-CoA panel by isothermal titration calorimetry against full-length ACBD3 and the isolated ACB domain (residues 83-174), then isolate oligomerisation-deficient ACB point mutants that retain ligand binding and ask whether Golgi PI4KB recruitment, PI4P levels, PKA RII anchoring or FAPP2-dependent glucosylceramide transport change. A ligand-binding-null mutant tested in parallel separates the ligand from the oligomer. If none of the Golgi readouts move, acyl-CoA binding is a genuinely separate arm of the protein and the SREBP1/lipogenesis axis is where to look for its consequence.
Hypothesis: Acyl-CoA-induced oligomerisation of the ACB domain gates ACBD3's scaffolding output rather than being incidental to it.
Type: biophysics
Experiment: Compare PI4KB activity when recruited to membranes by ACBD3 against PI4KB artificially tethered to the same membranes at matched surface density, to separate the local-concentration effect from any conformational activation.
Hypothesis: ACBD3 stimulates PI4KB purely by membrane recruitment rather than by allosteric activation.
Type: in vitro reconstitution
Experiment: Image ACBD3 against cis- and trans-Golgi markers at Airyscan resolution in SCFD1-knockout cells and in giantin/golgin-45 double-knockout cells, and ask whether the MWT374-376>AAA mutant retains any SCFD1-dependent residual pool. The reported model is sequential, with SCFD1 upstream, but the mutant and the UR-less truncation bind SCFD1 and SEC22B more strongly than wild type, so the two mechanisms may instead serve different subcompartments - which the sequential model would need to accommodate.
Hypothesis: The SCFD1-SEC22B step and the golgin step deliver ACBD3 to different Golgi subcompartments rather than acting strictly in series.
Type: cell biology
Experiment: The premise is the 2024 paper's own: the 3A peptide must outcompete the endogenous Golgi-localised ACBD3 recruitment factor, and its Figure 1D places the 3A contact residues (I380/K381) in the same unique region as MWT374-376. Independently, PMID:37044218 places the RIIalpha-binding interface in the single alpha helix of the GOLD domain, roughly 379-383: K381P abolishes binding but is a helix-breaking proline, Q379P and I380P have almost no effect, and Ile380 is proposed as the contact only from inspecting the apo structure. So the PKA and viral sites overlap as surfaces across residues 374-381, and whether they share a contact residue is precisely what is untested. Compete purified 3A against giantin and golgin-45 fragments, and against PKA RIIalpha, using non-proline substitutions across the 379-383 helix as separation-of-function probes - charge-reversal or bulk changes at I380 rather than I380P, which reported almost no effect and so cannot distinguish a non-contact from a tolerated substitution - for the ACBD3 328-528 fragment by quantitative pull-down or SPR, with PI4KB binding to the Q domain measured in parallel as the internal negative control. In infected cells, test whether the giantin-bound and RII-bound pools of Golgi ACBD3 fall while Golgi PI4KB is retained, and whether KDEL-receptor retrograde transport becomes constitutive as it does on ACBD3 depletion. A positive result would make the hijack a displacement of ACBD3's own anchors rather than occupancy of a spare site, and would predict a specific collateral consequence for host PKA signalling.
Hypothesis: Picornaviral 3A hijacks ACBD3 by outcompeting its endogenous anchors - the golgins at MWT374-376 and PKA RII at the 379-383 helix - all of which engage the unique region/GOLD surface, while PI4KB on the Q domain is untouched.
Type: biochemistry
Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)