GP is the envelope glycoprotein of Zaire ebolavirus and the only viral protein displayed on the virion surface. It is synthesised as a signal-peptide-bearing, heavily N- and O-glycosylated type I membrane precursor from the +1A edited GP mRNA, and is cleaved by host furin in the trans-Golgi into the receptor-binding subunit GP1 and the membrane-anchored fusion subunit GP2, which remain disulfide-linked and assemble into chalice-shaped homotrimeric spikes (GP1,2). GP1 carries the receptor-binding region and a large, disordered mucin-like domain; attachment to target cells is assisted by host lectins (CD209/DC-SIGN, CLEC4M, CLEC10A) and integrins, after which virions are internalised mainly by clathrin-mediated endocytosis. In the acidified endosome, cathepsins B and L strip the glycan cap from GP1, exposing the site that binds the entry receptor NPC1; NPC1 engagement together with low pH and calcium triggers the class I fusion rearrangement of GP2, whose fusion peptide inserts into the endosomal membrane and drives fusion, delivering the nucleocapsid to the cytoplasm. In infected cells GP concentrates in cholesterol-dependent plasma-membrane lipid rafts, the assembly and budding site, where it is incorporated into filamentous virions and where it antagonises the restriction factor BST2/tetherin to permit virion release. GP also has pathogenic activities beyond entry: it down-modulates surface integrins and MHC class I, causing cell rounding, detachment and anoikis of endothelial cells; it engages host TLR4 to kill bystander CD4+ T lymphocytes and to differentiate monocytes; and host ADAM17 sheds a soluble GP1,2-delta ectodomain complex (shed GP) that circulates in infected animals, acts as a decoy for neutralising antibodies, and activates dendritic cells and macrophages through TLR4 while increasing vascular permeability.
Curated functional classes representing distinct biological activities. These may be splice variants, cleavage products, or other forms with different functions.
GP_SHED_GPPRO_0000445686 (residues 33-637) GP_GP1PRO_0000037486 (residues 33-501) GP_GP2PRO_0000037487 (residues 502-676) | GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0001907 symbiont-mediated killing of host cell | EXP PMID:15051379 Ebola virus glycoprotein-mediated anoikis of primary human c... | ACCEPT | Summary: Adenovirus-delivered GP in primary human cardiac microvascular endothelial cells caused loss of adhesion followed by apoptotic death, and conditioned medium did not transfer the effect, so the killing requires intracellular GP expression rather than a secreted factor. The term is correct for this phenotype. Reason: Killing of host cells is one of GP's documented activities and not merely a side effect of overexpression: the cytotoxicity maps to the GP1 mucin-like region, has been seen in several cell types and delivery systems, and is a recognised contributor to the loss of vascular integrity in Ebola virus disease, so it is retained as part of the immune- and host-subversion arm of this review's core functions. Two caveats are recorded rather than acted on: delivery here is adenoviral overexpression, and UniProt marks the inference from cell detachment to haemorrhage as Probable. Kept at this granularity because the assay scores death of the GP-expressing host cell, which is exactly what the term asserts, and the mechanism proceeds through down-modulation of host integrins and MHC class I. Supporting Evidence: PMID:15051379 Primary human cardiac microvascular endothelial cells (HCMECs) transduced with Ad-EGP displayed loss of cell adhesion from the plastic surface followed by cell death. PMID:15051379 suggesting that EGP needs to be expressed intracellularly to exert its cytotoxic effect |
| GO:0001907 symbiont-mediated killing of host cell | EXP PMID:28542576 Ebola virus glycoprotein directly triggers T lymphocyte deat... | ACCEPT | Summary: GP added to isolated primary human CD4+ T cells killed them without infecting them; immunoprecipitation and flow cytometry showed GP binds the T cells through TLR4, and the TLR4 inhibitor CLI-095 rescued viability. This is direct, well-controlled support for GP-mediated killing of a host cell, here a bystander cell. Reason: A mechanistically defined activity of GP itself: GP-TLR4 engagement causes apoptosis and necrosis of uninfected lymphocytes, the likely basis of the lymphopenia that correlates with fatal outcome, and UniProt curates it with experimental evidence. GP performs the binding, so the claim is about this gene product rather than about a host factor. The term covers both the cell-autonomous and the bystander mode, so no more specific term is needed. Supporting Evidence: PMID:28542576 we demonstrate that EBOV directly binds to CD4+ T cells through interaction of GP with TLR4 PMID:28542576 The addition of the TLR4 inhibitor CLI-095 significantly reduced CD4+ T cell death induced by GP. |
| GO:0005576 extracellular region | EXP PMID:15103332 Ectodomain shedding of the glycoprotein GP of Ebola virus. | ACCEPT | Summary: Ectodomain shedding by host ADAM17/TACE at D637 removes the transmembrane anchor and liberates soluble GP1 plus truncated GP2 complexes from the cell surface; shed GP was detected in large amounts in the blood of infected animals. The extracellular location of this cleavage product is established. Reason: Direct experimental demonstration that a GP-derived product is released into the extracellular milieu, and the location is functionally meaningful: circulating shed GP is the antibody decoy and the TLR4 agonist. Strictly the claim is about the Shed GP chain (PRO_0000445686, residues 33-637) rather than the membrane-anchored spike, which is recorded under functional_isoforms rather than as an isoform qualifier on this row, since GOA carries no isoform field for it. Supporting Evidence: PMID:15103332 Proteolytic cleavage taking place at amino-acid position D637 removes the transmembrane anchor and liberates complexes consisting of GP1 and truncated GP2 (GP(2delta)) subunits from the cell surface. PMID:15103332 shed GP is present in significant amounts in the blood of virus-infected animals |
| GO:0005576 extracellular region | EXP PMID:25412102 Shed GP of Ebola virus triggers immune activation and increa... | ACCEPT | Summary: Shed GP released during infection binds and activates non-infected dendritic cells and macrophages, inducing cytokine secretion and increased vascular permeability, which requires it to act in the extracellular space. Reason: Independent experimental support for the same claim as the PMID:15103332 row, from a study whose whole design depends on soluble, cell-free GP. Reinforces rather than duplicates that row. Supporting Evidence: PMID:25412102 In effect it binds and activates non-infected dendritic cells and macrophages inducing the secretion of pro- and anti-inflammatory cytokines |
| GO:0005576 extracellular region | EXP PMID:30085081 Ebola Virus Shed Glycoprotein Triggers Differentiation, Infe... | ACCEPT | Summary: Treatment of monocytic cells with soluble shed GP drove their differentiation, infection and death through TLR4, again assaying GP as an extracellular agonist. Reason: Third independent experimental line for the extracellular localisation of shed GP; also shows the location is where a distinct function is exercised. Supporting Evidence: PMID:30085081 treatment of monocytic cells with Ebola virus shed glycoprotein (GP) promotes their differentiation resulting in increased infection and cell death |
| GO:0005576 extracellular region | IEA GO_REF:0000044 | ACCEPT | Summary: Keyword and subcellular-location mapping of the curated Secreted location of the Shed GP chain. The underlying UniProt statement rests on three experimental publications, all of which are reviewed above. Reason: An electronic mapping whose source statement is itself experimentally grounded, and which agrees with the three EXP rows for the same term. Kept for consistency with those rows. Supporting Evidence: file:EBOZM/GP/GP-uniprot.txt SUBCELLULAR LOCATION: [Shed GP]: Secreted |
| GO:0008289 lipid binding | EXP PMID:17545161 Structure of the Ebola fusion peptide in a membrane-mimetic ... | ACCEPT | Summary: The GP2 internal fusion peptide EBO16 folds into a 3(10)-helix in membrane-mimetic environments and interacts with detergent-resistant membrane fractions, with Trp8 and Phe12 required to maintain the structure inside the bilayer. Membrane and lipid engagement by the fusion loop is the physical basis of GP-mediated fusion. Reason: The claim is correct and mechanistically central: the fusion peptide must insert into the target (endosomal) membrane for GP2 to fuse membranes, and UniProt annotates the fusion peptide as a sequence feature. The term is kept at this granularity because the study used SDS micelles and bulk detergent-resistant membranes and therefore identifies no lipid class; a more specific child such as a particular phospholipid binding term would overstate the evidence. Supporting Evidence: PMID:17545161 EBO(16) can interact with detergent-resistant membrane fractions file:EBOZM/GP/GP-uniprot.txt Low pH in endosomes induces an irreversible conformational change in GP2, releasing the fusion hydrophobic peptide. |
| GO:0008289 lipid binding | IDA PMID:17545161 Structure of the Ebola fusion peptide in a membrane-mimetic ... | ACCEPT | Summary: Same experimental basis as the EXP row above, coded as a direct assay (NMR and tryptophan-fluorescence monitoring of peptide insertion into micelles and detergent-resistant membranes). Reason: Duplicate coding of one experiment rather than independent evidence, but the claim is sound and REMOVE is never the way to deduplicate rows that share a term. Action kept identical to the other lipid binding rows. Supporting Evidence: PMID:17545161 EBO(16) can interact with detergent-resistant membrane fractions |
| GO:0008289 lipid binding | IEA GO_REF:0000117 | ACCEPT | Summary: ARBA association-rule prediction of lipid binding for filovirus glycoproteins. The prediction is unvetted in itself, but it coincides with the experimental fusion-peptide data for this protein. Reason: The claim is true for GP on experimental grounds (see the PMID:17545161 rows), so the row is retained with the same action; its value is only confirmatory, since an ARBA rule is a machine-generated association and is not evidence in its own right. |
| GO:0008289 lipid binding | IPI PMID:17545161 Structure of the Ebola fusion peptide in a membrane-mimetic ... | ACCEPT | Summary: Third coding of the same fusion-peptide study, as a physical interaction with ChEBI:8984 (sodium dodecyl sulfate). The lipid-binding claim stands on the detergent-resistant-membrane data, but the WITH/FROM entity is a membrane-mimetic detergent, not a lipid, so it does not name a physiological ligand. Reason: Kept, consistently with the other lipid binding rows, because the underlying claim is correct. The defect is in the WITH/FROM field rather than in the annotation: SDS micelles were the folding environment used to obtain the NMR structure, and treating them as the bound lipid is misleading. Recorded as a source defect for DisProt. Supporting Evidence: PMID:17545161 insertion of the peptide was modeled by experiments in which the tryptophan fluorescence and (1)H NMR were monitored in the presence of sodium dodecyl sulfate micelles or in the presence of detergent-resistant membrane fractions |
| GO:0019031 viral envelope | IEA GO_REF:0000043 | ACCEPT | Summary: Keyword mapping of Viral envelope protein. GP1,2 trimers are the spikes of the virion envelope and the only viral protein on the particle surface, so this is a core location. Reason: One of the best-established facts about this protein: the curated UniProt record states GP is the sole surface protein of the virion, and immunogold electron microscopy identifies the spikes on released particles as GP. Supporting Evidence: file:EBOZM/GP/GP-uniprot.txt Essential for infectivity, as it is the sole viral protein expressed at the virion surface. PMID:11877482 Immunogold staining of the VLPs with anti-Ebola GP antibodies demonstrated the identity of the spikes on the surface of the particles as Ebola GP |
| GO:0019062 virion attachment to host cell | IEA GO_REF:0000043 | TOO GENERAL | Summary: Keyword mapping of Viral attachment to host cell. The claim is true, but this review also retains GO:0098670 entry receptor-mediated virion attachment to host cell, a descendant that states the same thing more precisely (GP1 binding to NPC1 and to host lectins). Reason: The term is correct for GP and is not being questioned; it is uninformative here because the more specific attachment term asserted elsewhere in this review entails it. This judges the term, not the keyword-based evidence: UniProt carries both keywords, and the specific one is the one worth keeping. |
| GO:0020002 host cell plasma membrane | IEA GO_REF:0000044 | ACCEPT | Summary: Mapping of the curated Host cell membrane location of GP1 and GP2. In infected and in transfected cells GP reaches the plasma membrane, where it colocalises with the raft marker GM1 in large surface patches and from where virions and virus-like particles bud. Reason: The electronic mapping rests on a curated location statement that is in turn based on confocal imaging of intact cells and on the budding electron microscopy of PMID:11877482. This is where GP is incorporated into virions and where it antagonises BST2, so the location is functionally core. Supporting Evidence: PMID:11877482 a substantial portion of both of the GPs were found to colocalize with GM1 in large patches on the plasma membrane file:EBOZM/GP/GP-uniprot.txt In the cell, localizes to the plasma membrane lipid rafts, which probably represent the assembly and budding site. |
| GO:0030430 host cell cytoplasm | IDA PMID:20202662 Ebola virus uses clathrin-mediated endocytosis as an entry p... | TOO GENERAL | Summary: CACAO row. PMID:20202662 is an entry-pathway study: it asks whether EbGP-pseudotyped HIV and replication-competent ZEBOV require clathrin-mediated endocytosis, using chlorpromazine, hypertonic sucrose, dominant-negative Eps15 and clathrin heavy-chain siRNA, with bafilomycin A1 for pH dependence. Having read the complete text (Materials and methods plus Results, recovered from PubMed Central; the cached copy holds only the abstract, introduction and discussion), the paper contains no assay of where GP is located. Every imaging experiment in it visualises host trafficking machinery, eGFP-clathrin, labelled transferrin, transferrin receptor and mRFP-Eps15, and the viral readout throughout is GFP reporter infectivity measured by flow cytometry or by counting infected cells. The cited paper therefore does not support an IDA cellular-component annotation on the viral glycoprotein. Independently, the term is also entailed by the host cell endoplasmic reticulum row kept below, since GO:0044165 is part_of GO:0030430. Reason: Two separate problems point to the same disposition. First, the claim as such is true at this altitude but says nothing useful: GP transits the host secretory pathway, and the endoplasmic reticulum annotation retained in this review already entails host cell cytoplasm, so this row adds no information. Second, and this is the CACAO-specific point, the cited reference does not show it: the row looks like a mechanism paper read as a localisation, with sentences about clathrin being redistributed in the cytoplasm, or about the virus entering the cytoplasm, converted into a component annotation on GP. It would in any case be a poor location for a single-pass type I membrane glycoprotein whose cytoplasmic tail is five residues long and whose ectodomain faces the lumen or the outside of every membrane it occupies. TOO_GENERAL rather than REMOVE because the term is genuinely entailed by a location this review keeps; the reference problem is recorded as a source defect instead. Supporting Evidence: PMID:20202662 Sensitivity of EbGP pseudotyped HIV as well as infectious EBOV to inhibitors of clathrin suggests that EBOV uses clathrin-mediated endocytosis as an entry pathway. |
| GO:0039587 symbiont-mediated-mediated suppression of host tetherin activity | IEA GO_REF:0000043 | KEEP AS NON CORE | Summary: Keyword mapping of Inhibition of host tetherin by virus. GP interacts with BST2 and counteracts its restriction of virion release; UniProt curates this from three experimental publications (PubMed:26516900, PubMed:27707924, PubMed:29669839). Reason: A well-supported and appropriately specific immune-evasion activity, but a distinct accessory function of GP rather than the receptor-binding and fusion activity the protein exists to perform. The keyword rests on curated statements with experimental evidence rather than on prediction, so it is kept rather than marked over-annotated. Supporting Evidence: file:EBOZM/GP/GP-uniprot.txt Interacts with BST2; this interaction inhibits the antiviral effect of BST2 and this allows viral release from infected cells |
| GO:0039654 fusion of virus membrane with host endosome membrane | IEA GO_REF:0000043 | ACCEPT | Summary: Keyword mapping of Fusion of virus membrane with host endosomal membrane. This is GP2's defining activity and the compartment is right: entry is pH-dependent and requires endosomal cathepsin processing and NPC1 binding before GP2 can fuse. Reason: Core function. The endosomal location of the fusion event is independently supported by the bafilomycin A1 sensitivity of both EbGP pseudotypes and replication-competent ZEBOV in PMID:20202662, and by the requirement for endosomal cathepsin B/L cleavage of GP1 before NPC1 engagement. This is the most specific fusion term the evidence supports. Supporting Evidence: file:EBOZM/GP/GP-uniprot.txt NPC1-binding, Ca(2+) and acidic pH induce a conformational change of GP2, which unmasks its fusion peptide and permit membranes fusion PMID:20202662 Also, both chlorpromazine and Bafilomycin A1 inhibited entry of infectious EBOV. |
| GO:0039663 membrane fusion involved in viral entry into host cell | IEA GO_REF:0000043 | TOO GENERAL | Summary: Keyword mapping of Fusion of virus membrane with host membrane, the parent of the endosome-specific term accepted above. Reason: True of GP but subsumed: GO:0039654, accepted in this review, is a descendant of this term and names the membrane that is actually fused. The pair comes from two UniProt keywords at different granularities; only the specific one is informative. |
| GO:0042802 identical protein binding | IPI PMID:26917592 Structural and molecular basis for Ebola virus neutralizatio... | KEEP AS NON CORE | Summary: Self-interaction recorded by IntAct from the structural study of neutralising antibodies, which resolves GP1-GP2 heterodimers associating into a chalice-shaped homotrimer; antibody Fab100 even bridges two adjacent protomers. GP's functional unit is the trimeric spike, so homotypic self-association is real and not an artefact. Reason: Unlike bare protein binding, identical protein binding carries information here: it states that GP homo-oligomerises, which is required to build the spike. It is kept as non-core because the informative molecular functions are the receptor binding and membrane-fusion activities of the assembled trimer rather than the self-association itself, which is a structural prerequisite. Supporting Evidence: PMID:26917592 which associate to form a chalice-shaped trimer file:EBOZM/GP/GP-uniprot.txt Homotrimer; each monomer consists of a GP1 and a GP2 subunit linked by disulfide bonds |
| GO:0044165 host cell endoplasmic reticulum | IMP PMID:21223600 Full-length Ebola glycoprotein accumulates in the endoplasmi... | KEEP AS NON CORE | Summary: CACAO row, and the paper does assert exactly this: full-length GP transiently expressed in HEK293T cells accumulated in a compartment that colocalised with a DsRed2-ER marker next to the nuclear membrane (Figure 1B), whereas a mucin-deleted GP did not (Figure 2A). The mutant comparison is presumably why the row was coded IMP, although the observation on wild-type GP is a marker colocalisation, that is, an IDA. Endoplasmic reticulum residence is independently expected for a protein with a cleaved signal peptide, N- and O-glycosylation and furin processing in the trans-Golgi. Reason: The term matches what the paper shows and is biologically sound as a biosynthetic transit compartment, so it is kept; but three caveats keep it out of the core. The assay is transfection of 8 to 10 micrograms of plasmid for 24 hours with a protein that is itself cytotoxic, so accumulation may be an overexpression phenotype rather than the steady state during infection. The same study found no Golgi signal, which is hard to reconcile with the established furin cleavage of GP in the trans-Golgi and suggests the imaging did not capture the full itinerary. And the authors themselves note other work placing GP at the plasma membrane. GP does no work in the endoplasmic reticulum: it folds, oligomerises and is glycosylated there en route to the surface, so the functional locations are the host plasma membrane, the lipid raft and the virion membrane. Supporting Evidence: PMID:21223600 Co-expression of Ebola GP with DsRed2-ER showed that GP had localized within the ER (Figure 1B). PMID:21223600 it did not accumulate in the ER (Figure 2A) PMID:21223600 However, separate studies show that Ebola GP localizes in the plasma membrane of either tissues from experimentally infected non-human primates |
| GO:0045121 membrane raft | IDA PMID:11877482 Lipid raft microdomains: a gateway for compartmentalized tra... | ACCEPT | Summary: CACAO row that the cited paper supports squarely, and with unusual rigour for a single-reference cellular-component annotation. PMID:11877482 is a dedicated study of filovirus raft association: transfected Ebola GP partitioned into the raft fraction of sucrose gradients while a GM1 spot blot confirmed fraction quality (Fig. 1A and 1B); a GP mutated at the palmitoylation sites C670/C672 failed to reach rafts (Fig. 1B); cholesterol depletion with methyl-beta-cyclodextrin stripped GP out of the raft fraction and the raft-excluded transferrin receptor stayed soluble (Fig. 1C); and on intact cells GP colocalised with GM1 in large plasma-membrane patches by confocal microscopy (Fig. 2A). GP was also found in raft fractions of Ebola-infected cells, not only transfected cells. Reason: The CACAO term matches the assay exactly: it is neither a phenotype read as a process nor a single-image localisation, and it is internally controlled in both directions, by a mutant and a drug that abolish it and by a raft-excluded protein that does not partition with it. UniProt reaches the same conclusion from the same paper. The raft is the functionally relevant plasma-membrane subdomain for GP: it is the assembly and budding platform, and filovirus entry also requires intact rafts. Supporting Evidence: PMID:11877482 a significant fraction of Ebola and Marburg GPs were found to reside in rafts PMID:11877482 resulted in almost complete removal of Ebola GP from rafts PMID:11877482 a substantial portion of both of the GPs were found to colocalize with GM1 in large patches on the plasma membrane file:EBOZM/GP/GP-uniprot.txt Filoviruses entry requires functional lipid rafts at the host cell surface. |
| GO:0046718 symbiont entry into host cell | EXP PMID:17545161 Structure of the Ebola fusion peptide in a membrane-mimetic ... | ACCEPT | Summary: The GP2 fusion peptide study, read as evidence that GP mediates entry: the peptide that inserts into the host membrane and leads to fusion is an internal sequence of GP2. Reason: GP is the sole determinant of filovirus entry and this is its core process term. The cited paper supports the mechanism at the level of the fusion peptide; the whole-protein claim is supported far more broadly, including by the pseudotype experiments of PMID:20202662, in which EbGP alone confers entry on an envelope-minus HIV particle. Supporting Evidence: PMID:17545161 This region interacts with the cellular membrane of the host and leads to membrane fusion. file:EBOZM/GP/GP-uniprot.txt Trimeric GP1,2 complexes form the virion surface spikes and mediate the viral entry processes |
| GO:0046718 symbiont entry into host cell | IDA PMID:17545161 Structure of the Ebola fusion peptide in a membrane-mimetic ... | ACCEPT | Summary: Second coding of the same fusion-peptide study for the entry process. Reason: Same claim and same action as the EXP row from this reference. Duplicate coding of one experiment is not grounds for removal, and the claim itself is among the best-established facts about GP. Supporting Evidence: PMID:17545161 This region interacts with the cellular membrane of the host and leads to membrane fusion. |
| GO:0046718 symbiont entry into host cell | IEA GO_REF:0000120 | ACCEPT | Summary: Combined electronic annotation from an ARBA rule plus the Virus entry into host cell and Viral penetration into host cytoplasm keywords, all asserting the process that GP's experimental rows demonstrate. Reason: The electronic inference agrees with direct experimental evidence for this protein, so the row is retained with the same action as the experimental rows for the same term. |
| GO:0046718 symbiont entry into host cell | IMP PMID:20202662 Ebola virus uses clathrin-mediated endocytosis as an entry p... | ACCEPT | Summary: CACAO row, and the best-supported of the five. PMID:20202662 shows that an envelope-minus HIV particle bearing EbGP enters cells, and that this entry is abolished by four independent perturbations of the clathrin pathway (chlorpromazine, 0.45 M hypertonic sucrose, dominant-negative Eps15 DIII, clathrin heavy-chain siRNA) in HOS, HeLa, Vero, 293T and primary human endothelial cells, while wild-type HIV and VSVg-pseudotyped HIV are unaffected; chlorpromazine and bafilomycin A1 also block replication-competent ZEBOV-GFP. The pseudotype design makes GP the only viral determinant of entry. Reason: The CACAO term matches the paper's central result at the right granularity, and this is GP's core biological process. The IMP coding is appropriate in that the perturbed entities are host factors whose loss abolishes GP-mediated entry; the claim about GP is independently established by the pseudotype assay itself. Note that the same reference also carries a GO:0030430 row, which this review does not keep at face value: this row, not that one, is what the paper shows. Supporting Evidence: PMID:20202662 Inhibition of clathrin pathway using dominant-negative Eps15, siRNA-mediated knockdown of clathrin heavy chain, chlorpromazine and sucrose blocked EbGP pseudotyped HIV infection. PMID:20202662 Sensitivity of EbGP pseudotyped HIV as well as infectious EBOV to inhibitors of clathrin suggests that EBOV uses clathrin-mediated endocytosis as an entry pathway. |
| GO:0046761 viral budding from plasma membrane | IDA PMID:11877482 Lipid raft microdomains: a gateway for compartmentalized tra... | KEEP AS NON CORE | Summary: CACAO row supported by the paper's second half. Transfected GP alone was released into the supernatant as particulate material that immunoprecipitated with the raft marker GM1, while the raft-deficient GP C670/672A mutant was not; coexpression of GP with VP40 drove the majority of both proteins into the supernatant; the released particles were filamentous, carried GP spikes identified by immunogold, and contained GM1 but excluded transferrin receptor. Pre-embedment immunogold electron microscopy showed GP-positive particles exiting through discrete regions of the plasma membrane. Reason: The term and the compartment match the data, so the row is kept. It is non-core because VP40 is the filovirus budding driver, and GP's contribution is to be incorporated into the particle and to raise release efficiency from the raft platform, helped by its antagonism of BST2/tetherin, rather than to drive membrane scission itself. Granularity is right: the plasma membrane, not an internal compartment, is where the electron microscopy shows release occurring. Supporting Evidence: PMID:11877482 coexpression of GPwt and VP40, directed the majority of both proteins into the supernatant PMID:11877482 where a large number of particles that stain for GP exit through a small region of the plasma membrane file:EBOZM/GP/GP-uniprot.txt In the cell, localizes to the plasma membrane lipid rafts, which probably represent the assembly and budding site. |
| GO:0052031 symbiont-mediated perturbation of host defense response | IEA GO_REF:0000043 | TOO GENERAL | Summary: Keyword mapping of Viral immunoevasion. GP certainly perturbs host defence, but this review keeps two descendants of this term that say how: suppression of the innate immune response, and specifically of tetherin activity. Reason: Correct but subsumed by the more specific immune-evasion terms retained in this review, both of which are descendants of it. No doubt is being cast on the claim or on the keyword. |
| GO:0052170 symbiont-mediated suppression of host innate immune response | IEA GO_REF:0000043 | ACCEPT | Summary: Keyword mapping of Inhibition of host innate immune response by virus. GP blunts innate defence in several documented ways: shed GP absorbs neutralising antibodies and diverts dendritic-cell and macrophage activation, GP antagonises the restriction factor BST2, and GP down-regulates surface molecules required for immune surveillance. Reason: Well-grounded in curated experimental statements and appropriately specific for the several mechanisms involved. Immune subversion is treated here as a core function alongside entry rather than as a peripheral one: shedding of GP, the TLR4 agonism of the shed and membrane forms, and the down-regulation of surface immune molecules are conserved, dedicated activities of this gene product, not incidental consequences of entry. Kept at this level because the individual mechanisms are heterogeneous and only the tetherin arm has a dedicated child term. Supporting Evidence: file:EBOZM/GP/GP-uniprot.txt At later times of infection, down-regulates the expression of various host cell surface molecules that are essential for immune surveillance and cell adhesion PMID:15103332 by efficiently blocking the activity of virus-neutralizing antibodies |
| GO:0055036 virion membrane | IEA GO_REF:0000044 | ACCEPT | Summary: Mapping of the curated Virion membrane location of GP1 and GP2. GP2 is a single-pass type I membrane protein of the envelope and GP1 is held on the particle as a disulfide-linked peripheral partner. Reason: Core location: this is where the receptor-binding and fusion activities are exercised. The curated statement is supported experimentally, and purified virions and virus-like particles carry GP spikes detectable by immunogold electron microscopy. Supporting Evidence: file:EBOZM/GP/GP-uniprot.txt GP1 is not anchored to the viral envelope, but forms a disulfid-linked complex with the extravirion surface GP2 |
| GO:0075509 endocytosis involved in viral entry into host cell | IEA GO_REF:0000043 | TOO GENERAL | Summary: Keyword mapping of Virus endocytosis by host, the parent of the clathrin-specific term accepted below. Reason: True of GP, but GO:0075512 clathrin-dependent endocytosis of virus by host cell is retained in this review and is a descendant that names the route actually demonstrated for Ebola virus. The pair is an artefact of two keywords at different granularities. |
| GO:0075512 clathrin-dependent endocytosis of virus by host cell | IEA GO_REF:0000043 | ACCEPT | Summary: Keyword mapping of Clathrin-mediated endocytosis of virus by host. Unusually for an IEA, the specific claim has been demonstrated directly for this protein: EbGP-mediated infection is blocked by chlorpromazine, hypertonic sucrose, dominant-negative Eps15 and clathrin heavy-chain knockdown, with wild-type HIV and VSVg controls unaffected, and chlorpromazine also blocks replication-competent Ebola virus. Reason: The electronic row coincides with direct experimental evidence reviewed above (PMID:20202662), so it is accepted at this specificity. The authors note Ebola virus may use additional routes such as macropinocytosis in some settings, which does not weaken the clathrin claim. Supporting Evidence: PMID:20202662 Inhibition of clathrin pathway using dominant-negative Eps15, siRNA-mediated knockdown of clathrin heavy chain, chlorpromazine and sucrose blocked EbGP pseudotyped HIV infection. |
| GO:0098670 entry receptor-mediated virion attachment to host cell | IEA GO_REF:0000043 | ACCEPT | Summary: Keyword mapping of Viral attachment to host entry receptor. GP1 is the receptor-binding subunit; after cathepsin B/L removal of the glycan cap it binds the entry receptor NPC1, and this binding together with low pH and calcium triggers fusion. Reason: Core function at the right granularity: the claim is about engagement of the entry receptor proper (NPC1), not merely of attachment factors, and UniProt curates the NPC1 interaction from experimental publications. This is the term that makes the more general GO:0019062 row redundant. Supporting Evidence: file:EBOZM/GP/GP-uniprot.txt This cleaved 19-kDa GP1 can then bind to the host entry receptor NPC1 file:EBOZM/GP/GP-uniprot.txt SUBUNIT: [GP1]: Interacts with host entry receptor NPC1. |
| GO:0046789 host cell surface receptor binding | IPI PMID:26917592 Structural and molecular basis for Ebola virus neutralizatio... | NEW | Summary: Proposed new molecular-function annotation. GOA gives GP no molecular function for the activity it is famous for: binding host receptors. GP1 is the receptor-binding subunit, and UniProt curates experimentally determined interactions with the entry receptor NPC1 and with the host surface molecules CD209/DC-SIGN, CLEC4M, CLEC10A, FOLR1, ITGAV and FCN1. Reason: GP itself performs the binding, so the function belongs on this gene product rather than on a partner; this is not a necessity-versus-participation case. The comparator check supports it: GO:0046789 is the standard molecular function for class I viral fusion glycoproteins that engage host receptors, and GP's existing annotations capture the process (entry receptor-mediated virion attachment, symbiont entry) while leaving the molecular function empty. The term is neither an ancestor nor a descendant of any term GP already carries. Conservatively proposed at the parent level rather than inventing a receptor-specific child, and the evidence is a physical interaction with host proteins rather than a phenotype. Supporting Evidence: PMID:26917592 The GP1 subunit binds to the EBOV receptor, Niemann-Pick C1 (NPC1), allowing GP2-mediated fusion of the viral and host-cell membranes file:EBOZM/GP/GP-uniprot.txt Responsible for binding to the receptor(s) on target cells. file:EBOZM/GP/GP-uniprot.txt SUBUNIT: [GP1]: Interacts with host entry receptor NPC1. |
| GO:0035662 Toll-like receptor 4 binding | IPI PMID:28542576 Ebola virus glycoprotein directly triggers T lymphocyte deat... | NEW | Summary: Proposed new molecular-function annotation. Both membrane GP and shed GP act on uninfected immune cells through host TLR4, and the interaction has been shown directly by immunoprecipitation and flow cytometry on primary CD4+ T cells, with the TLR4 inhibitor CLI-095 abolishing the downstream death. UniProt curates the interaction with experimental evidence from two publications. Reason: This is the molecular function underlying two claims GOA already carries as processes (symbiont-mediated killing of host cell from PMID:28542576, and the extracellular shed-GP rows), and GP is the entity that does the binding. The comparator check is favourable: GO:0035662 is annotated to the host ligands and accessory proteins of TLR4, and annotating a microbial agonist that binds it directly follows the same pattern. The term is not an ancestor or descendant of anything GP already carries. Evidence is a physical interaction assay, not a knockout phenotype. Supporting Evidence: PMID:28542576 Using immunoprecipitation assays and flow cytometry, we demonstrate that EBOV directly binds to CD4+ T cells through interaction of GP with TLR4. PMID:30085081 The effects were inhibited by blocking Toll-like receptor 4 pathway. file:EBOZM/GP/GP-uniprot.txt Interacts with host TLR4; this interaction induces cell death in T-lymphocytes or proinflammatory cytokines and SOCS1 production in monocytes |
Loading supporting content…
Download this section (compressed HTML)Q: Does the raft dependence of GP assembly and budding reflect a requirement for a specific lipid environment at the budding site, or simply colocalisation with VP40, given that palmitoylation-site mutants lose raft partitioning but are reported not to lose function?
Suggested experts: Bavari S, Aman MJ
Q: Is the endoplasmic reticulum accumulation of full-length GP seen on plasmid overexpression detectable in Ebola-virus-infected cells, and does it contribute to GP-induced cell rounding?
Suggested experts: Bhattacharyya S, Hope TJ
Q: Shed GP and membrane GP both act through TLR4. Are the relevant epitopes the same, and does the mucin-like region or its glycans constitute the TLR4-engaging surface?
Suggested experts: Bukreyev A, Volchkov VE
Experiment: Measure binding and insertion of recombinant GP2 ectodomain and of the EBO16 peptide into liposomes of defined composition (varying cholesterol, the late-endosomal lipid bis(monoacylglycero)phosphate, phosphatidylserine and sphingomyelin) by surface-plasmon resonance and tryptophan-fluorescence depth quenching, and test whether the same compositions support NPC1- and low-pH-triggered fusion of GP-bearing pseudovirions in a lipid-mixing assay.
Hypothesis: The GP2 fusion loop binds a defined lipid class rather than bulk hydrophobic membrane, and that preference sets the endosomal membrane requirement for fusion.
Type: liposome binding and in vitro fusion
Experiment: Compare GP localisation in cells infected with recombinant Ebola virus at a low multiplicity against cells transfected with GP plasmid at matched expression levels, using quantitative immunofluorescence against markers for endoplasmic reticulum, ERGIC, trans-Golgi and plasma membrane, with a pulse-chase of endoglycosidase-H-sensitive versus resistant GP to measure the rate of Golgi transit.
Hypothesis: GP localises to the endoplasmic reticulum during authentic infection only transiently, and the apparent accumulation reported from plasmid transfection is an overexpression artefact.
Type: infection versus transfection localisation and glycan maturation time course
Experiment: Build a recombinant Ebola virus carrying GP C670A/C672A, with a single-site control, by reverse genetics, then quantify GP incorporation per particle by western blot normalised to VP40, particle morphology by electron microscopy, and specific infectivity per genome copy; test whether cholesterol depletion of producer cells phenocopies the mutant.
Hypothesis: Raft targeting via GP2 palmitoylation is required for efficient incorporation of GP into budding virions and therefore for particle infectivity.
Type: reverse genetics with particle composition and specific infectivity measurement
What is not known — curated, literature-grounded statements of the open unknowns (the inverse of core functions).
Gap: The lipid species that the GP2 internal fusion loop engages in the endosomal membrane is unidentified, so the chemistry of the insertion step is unknown.
BIOLOGY MF_DARK
What is known: The fusion peptide (residues 524-539) is mapped, its 3(10)-helical fold in membrane-mimetic environments is solved, Trp8 and Phe12 are known to be required for the fold, and the peptide interacts with bulk detergent-resistant membrane fractions. But the published assays used SDS micelles and whole detergent-resistant membranes, and the GOA WITH/FROM entity for the lipid-binding annotation is sodium dodecyl sulfate, a detergent.
Significance: Fusion-loop lipid preference determines which endosomal membrane composition supports fusion and is the target for small molecules that block the fusion step rather than receptor binding.
Gap: Whether full-length GP accumulates in the endoplasmic reticulum during authentic infection, or only when overexpressed from plasmids, is undetermined, and so is the mechanism by which the mucin-like region causes the accumulation.
BIOLOGY CC_DARK
What is known: Transiently overexpressed full-length GP colocalises with an endoplasmic reticulum marker in HEK293T cells at 24 hours while a mucin-deleted GP does not, and the mucin-like region is independently known to carry the cytotoxic activity. But GP also reaches the plasma membrane in infected non-human-primate tissue, and must pass through the trans-Golgi to be cleaved by furin, a compartment in which the same study saw no signal.
Significance: If the accumulation is real in infection, retention of a cytotoxic glycoprotein in the endoplasmic reticulum would be a candidate mechanism for GP-induced cell rounding and anoikis; if it is an artefact of overexpression, the cytotoxicity must be explained at the cell surface.
Gap: It is not known whether GP2 palmitoylation at Cys670 and Cys672 is required for infectivity, despite being required for lipid-raft partitioning.
BIOLOGY MF_DARK
What is known: The C670/672A mutant fails to enter rafts and fails to release GM1-containing particles, yet UniProt records that palmitoylation is not required for GP2 function on the basis of a separate study.
Significance: Resolving the conflict would establish whether raft targeting is mechanistically necessary for assembly and budding or merely correlated with it.
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)