NDUFAF3 (also known as C3orf60) is a nucleus-encoded, mitochondrially imported assembly factor for mitochondrial respiratory chain complex I (NADH:ubiquinone oxidoreductase). It is not a structural subunit of the mature holoenzyme and has no catalytic activity. NDUFAF3 associates with the mitochondrial inner membrane and acts early in complex I biogenesis, tightly interacting with its partner factor NDUFAF4 and with the iron-sulfur (IP) core subunits NDUFS2 and NDUFS3 to help nucleate and stabilize an early membrane-arm assembly intermediate. Like other complex I assembly factors, it associates transiently with assembly intermediates and dissociates from the fully assembled complex. Biallelic loss-of-function variants in NDUFAF3 cause isolated mitochondrial complex I deficiency (nuclear type 18), with phenotypes ranging from fatal neonatal mitochondrial disease to Leigh syndrome.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0032981 mitochondrial respiratory chain complex I assembly | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic (IBA) propagation of the core biological process for this gene. This is the well-supported, evolutionarily conserved role of NDUFAF3 as a complex I assembly factor, directly demonstrated experimentally in human by the disease-causing loss-of-function study (PMID:19463981). Accept as core. Reason: The PAINT node placing complex I assembly on this family is supported by direct experimental work on the human gene itself: patient loss-of-function mutations cause isolated complex I deficiency and NDUFAF3-GFP complementation restored it. NDUFAF3 appearing among the descendants that grounded the node is expected, not circular. No evidence of lineage-specific loss argues against the placement. Supporting Evidence: PMID:19463981 We found that NDUFAF3 is a genuine mitochondrial complex I assembly protein that interacts with complex I subunits. |
| GO:0005743 mitochondrial inner membrane | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic (IBA) propagation of the mitochondrial inner membrane localization. Consistent with UniProt subcellular location and with the experimental IDA localization in PMID:19463981. Accept. Reason: The inner membrane is where complex I is built, and the human protein was localized there experimentally in the primary study, so the inherited assertion is confirmed rather than merely tolerated. `is_active_in` is the right qualifier for a transient assembly factor that acts at, but is not retained in, the membrane-embedded complex. Supporting Evidence: file:human/NDUFAF3/NDUFAF3-uniprot.txt Nucleus {ECO:0000250}. Mitochondrion inner membrane {ECO:0000269|PubMed:19463981}. |
| GO:0005634 nucleus | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: Electronic (IEA) mapping from the UniProt Subcellular Location keyword to the nucleus GO term. UniProt records a Nucleus location only by similarity (ECO:0000250), not from direct evidence on this protein, and the functional role of NDUFAF3 is entirely mitochondrial (complex I assembly). The nuclear localization is unverified for this protein and not part of its core function; keep as non-core rather than removing, since it derives from the UniProt annotation. Reason: The keyword mapping faithfully reproduces what UniProt records, but what UniProt records here is an ECO:0000250 by-similarity assertion, not an observation on this protein; the only experimentally evidenced location in the entry is the inner membrane. A nuclear pool with no demonstrated function is not part of the core annotation, but the term is not contradicted either, so it is demoted rather than removed. Supporting Evidence: file:human/NDUFAF3/NDUFAF3-uniprot.txt Nucleus {ECO:0000250}. Mitochondrion inner membrane {ECO:0000269|PubMed:19463981}. |
| GO:0005743 mitochondrial inner membrane | IEA GO_REF:0000044 | ACCEPT | Summary: Electronic (IEA) mapping from the UniProt Subcellular Location keyword to the mitochondrial inner membrane GO term. This is the experimentally supported location of NDUFAF3 (IDA in PMID:19463981) and where it functions as an assembly factor. Accept. Reason: Unlike the nucleus mapping from the same GO_REF, this one reproduces an experimentally evidenced UniProt location (ECO:0000269|PubMed:19463981) and matches the compartment in which NDUFAF3 does its work. Accepted. Supporting Evidence: file:human/NDUFAF3/NDUFAF3-uniprot.txt Nucleus {ECO:0000250}. Mitochondrion inner membrane {ECO:0000269|PubMed:19463981}. |
| GO:0032981 mitochondrial respiratory chain complex I assembly | IEA GO_REF:0000002 | ACCEPT | Summary: Electronic (InterPro2GO) inference of the complex I assembly process from the NDUF3 InterPro signature (IPR034095). This correctly captures the core function and is corroborated by direct experimental evidence. Accept as core. Reason: IPR034095 is the NDUFAF3-specific signature rather than a broad domain family, so the InterPro2GO mapping is narrowly targeted and yields exactly the process the human protein was experimentally shown to be required for. Accepted as corroborating the core process. Supporting Evidence: PMID:19463981 We found that NDUFAF3 is a genuine mitochondrial complex I assembly protein that interacts with complex I subunits. |
| GO:0005515 protein binding | IPI PMID:19688755 LC-MS/MS as an alternative for SDS-PAGE in blue native analy... | REMOVE | Summary: IPI protein binding annotation (with NDUFAF4/Q9P032). "protein binding" is an uninformative molecular function term. This interaction (NDUFAF3 with NDUFAF4) is biologically meaningful and central to NDUFAF3 function, but the generic GO:0005515 term does not convey it. Remove the uninformative term; removal does not dispute the reported interaction, which is retained in core_functions and in the SUBUNIT interactions. Reason: The evidence behind this row is protein correlation profiling across blue-native gel slices, which infers co-migration within a native complex rather than a direct pairwise contact; it is what first flagged C3ORF60/NDUFAF3 as a complex I biogenesis candidate. The underlying biology is real and is recorded in core_functions, but 'protein binding' communicates none of it, and co-migration evidence supports no more informative molecular function term to put in its place. Supporting Evidence: PMID:19688755 involved in the biogenesis of complex I, including the recently implicated |
| GO:0005515 protein binding | IPI PMID:24344204 TIMMDC1/C3orf1 functions as a membrane-embedded mitochondria... | REMOVE | Summary: IPI protein binding annotation (with NDUFAF4/Q9P032) from the TIMMDC1/MCIA interaction proteomics study. Uninformative generic MF term; the specific NDUFAF3-NDUFAF4 interaction is retained in core_functions. Remove the bare term. Reason: This row comes from an interaction-proteomics survey of 15 complex I core subunits and assembly factors that produced a 101-protein, 335-edge network in which NDUFAF3 sits as an MCIA-associated assembly factor. The network placement is informative; the bare GO:0005515 term is not, and network membership does not identify a specific molecular function to replace it with. The interaction itself is kept in core_functions and is not disputed. Supporting Evidence: PMID:24344204 resulting in a network of 101 proteins and 335 interactions |
| GO:0005515 protein binding | IPI PMID:25416956 A proteome-scale map of the human interactome network. | REMOVE | Summary: IPI protein binding from a proteome-scale binary interactome map (partners SNRPA/P09012 and RBCK1/Q9BYM8). These high-throughput Y2H interactions are not established mitochondrial/complex I biology and the generic MF term is uninformative. Remove. Reason: SNRPA is a spliceosomal snRNP protein and RBCK1 a cytoplasmic E3 ligase; neither is mitochondrial, and neither has any reported role in complex I biogenesis. These are systematic binary screen hits from a ~14,000-interaction map, not directed assays, so they support no specific molecular function for NDUFAF3 that could replace the generic term. Supporting Evidence: PMID:25416956 systematic map of ?14,000 high-quality human binary protein-protein interactions |
| GO:0005515 protein binding | IPI PMID:27499296 Mitochondrial Protein Interaction Mapping Identifies Regulat... | REMOVE | Summary: IPI protein binding annotation (with NDUFAF4/Q9P032) from a mitochondrial affinity-enrichment MS interaction map. Consistent with the biologically relevant NDUFAF3-NDUFAF4 partnership, but the bare GO:0005515 term is uninformative. Remove the term; the meaningful interaction is captured in core_functions. Reason: Affinity-enrichment MS on mitochondrial proteins of unclear function recovers the NDUFAF3-NDUFAF4 pair in its native compartment, which corroborates the partnership established in the primary study. It still licenses only the uninformative parent term and no more specific molecular function, so the row is removed while the interaction is recorded in core_functions. Supporting Evidence: PMID:27499296 interactions for 50 select MXPs using affinity enrichment mass spectrometry |
| GO:0005515 protein binding | IPI PMID:32296183 A reference map of the human binary protein interactome. | REMOVE | Summary: IPI protein binding from a proteome-scale binary interactome map, spanning many partners (e.g. AGTRAP, CYSRT1, GOLGA6L9, INCA1, KRT27, MBD3L1, POU2AF1, RNF146, SNRPA, UHMK1, and NDUFAF4). Most are high-throughput binary hits without mitochondrial context; the generic MF term is uninformative. Remove. Reason: Most of the listed partners (KRT27, GOLGA6L9, MBD3L1, POU2AF1 and the rest) are neither mitochondrial nor connected to complex I; they are all-by-all yeast two-hybrid hits, and the authors themselves note that the cellular function of most individual PPIs in the map remains unknown, so no specific molecular function can replace the generic term. The NDUFAF4 hit is meaningful but is already captured elsewhere. Supporting Evidence: PMID:32296183 the cellular function of most individual PPIs remains to be elucidated |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | REMOVE | Summary: IPI protein binding annotation (with NDUFAF4/Q9P032) from the BioPlex dual-proteome interactome. Consistent with the core NDUFAF3-NDUFAF4 interaction, but the bare GO:0005515 term is uninformative. Remove. Reason: BioPlex 3.0 recovers the NDUFAF3-NDUFAF4 interaction by affinity-purification MS in 293T cells, independently reproducing the core partnership. As with the other interactome sources the only supported GO term is the uninformative parent, so the row is removed and the biology retained in core_functions; the interaction is not in doubt. Supporting Evidence: PMID:33961781 of 10,128 human proteins-half the proteome-in 293T cells and includes 118,162 |
| GO:0005515 protein binding | IPI PMID:40205054 Multimodal cell maps as a foundation for structural and func... | REMOVE | Summary: IPI protein binding annotation (with NDUFAF4/Q9P032) from a multimodal cell-map interactome study. Recapitulates the core NDUFAF3-NDUFAF4 interaction, but the generic MF term is uninformative. Remove. Reason: A multimodal map (biophysical interactions plus immunofluorescence) of U2OS cells that resolves molecular assemblies rather than individual molecular functions; it recapitulates the NDUFAF3-NDUFAF4 pair but adds no functional resolution beyond the bare binding term, so there is no more specific term to substitute. Supporting Evidence: PMID:40205054 images for over 5,100 proteins in U2OS osteosarcoma cells |
| GO:0005739 mitochondrion | HTP PMID:34800366 Quantitative high-confidence human mitochondrial proteome an... | KEEP AS NON CORE | Summary: High-throughput (HTP) mitochondrial localization from a high-confidence quantitative mitochondrial proteome study. Correct but less specific than the experimentally supported mitochondrial inner membrane localization. Keep as a correct, non-core (more general) localization. Reason: Inclusion in MitoCoP, a quantitatively filtered high-confidence mitochondrial proteome, is sound evidence that NDUFAF3 resides in mitochondria, but an organelle-level proteomic inventory cannot resolve submitochondrial compartment. It therefore corroborates, at a coarser granularity, the IDA inner-membrane annotation that remains the annotation of record. Supporting Evidence: PMID:34800366 mitochondrial high-confidence proteome of >1,100 proteins (MitoCoP) |
| GO:0005743 mitochondrial inner membrane | TAS Reactome:R-HSA-6799178 | ACCEPT | Summary: Reactome TAS localization to the mitochondrial inner membrane, from the "Complex I biogenesis" pathway (step: Intermediate 1 binds HP subcomplex to form Intermediate 2). Consistent with the experimentally supported inner membrane location where NDUFAF3 acts. Accept. Reason: This Reactome step names NDUFAF3 explicitly as one of the Intermediate 1 assembly factors that anchor the hydrophobic protein (HP) fraction to the inner mitochondrial membrane. The membrane localization is therefore intrinsic to the role Reactome assigns it, and matches the experimental IDA. Accepted. Supporting Evidence: Reactome:R-HSA-6799178 anchored to the inner mitochondrial membrane by Intermediate |
| GO:0005743 mitochondrial inner membrane | TAS Reactome:R-HSA-6799179 | ACCEPT | Summary: Reactome TAS localization to the mitochondrial inner membrane, from a "Complex I biogenesis" assembly step. Consistent with the experimentally supported inner membrane location. Accept. Reason: A later step of the same Complex I biogenesis pathway, in which the peripheral-arm subunits join the 815 kDa intermediate at the inner membrane. Consistent with the experimental localization; accepted as pathway-level corroboration. Supporting Evidence: Reactome:R-HSA-6799179 Subunits NDUFA12, NDUFS1, 4, 6, NDUFV1, 2 and 3 with the assembly factor |
| GO:0005743 mitochondrial inner membrane | TAS Reactome:R-HSA-6799191 | ACCEPT | Summary: Reactome TAS localization to the mitochondrial inner membrane, from a "Complex I biogenesis" assembly step. Consistent with the experimentally supported inner membrane location. Accept. Reason: Reactome describes this step as the IP subcomplex being anchored to the inner mitochondrial membrane by MT-ND1, so the compartment assignment follows from the reaction itself rather than being asserted independently. Consistent with the experimental localization; accepted. Supporting Evidence: Reactome:R-HSA-6799191 to the inner mitochondrial membrane by NADH-ubiquinone oxidoreductase chain 1 |
| GO:0005743 mitochondrial inner membrane | TAS Reactome:R-HSA-6799196 | ACCEPT | Summary: Reactome TAS localization to the mitochondrial inner membrane, from the "Complex I biogenesis" step in which the MCIA complex and assembly factors (NDUFAF2-7, TIMMDC1) dissociate from the 980kDa complex to leave mature complex I. Consistent with the inner membrane location and with NDUFAF3 being a transient assembly factor that leaves the mature enzyme. Accept. Reason: The terminal step of the pathway, in which the MCIA complex and the assembly factors NDUFAF2-7 and TIMMDC1 - NDUFAF3 among them - dissociate from the 980 kDa intermediate to leave mature complex I. This both supports the inner-membrane localization and confirms that NDUFAF3 is transient rather than a holoenzyme subunit. Accepted. Supporting Evidence: Reactome:R-HSA-6799196 The MCIA complex, NDUFAF2-7 all dissociate from the 980kDa complex |
| GO:0005743 mitochondrial inner membrane | TAS Reactome:R-HSA-6799197 | ACCEPT | Summary: Reactome TAS localization to the mitochondrial inner membrane, from a "Complex I biogenesis" assembly step. Consistent with the experimentally supported inner membrane location. Accept. Reason: A membrane-arm step (MT-ND4 and MT-ND5 joining the 550 kDa complex) of the same inner-membrane pathway. Consistent with the experimental localization; accepted as pathway-level corroboration rather than independent evidence. Supporting Evidence: Reactome:R-HSA-6799197 Distal components of the membrane arm MT-ND4 and 5 associate with the 550kDa |
| GO:0005743 mitochondrial inner membrane | TAS Reactome:R-HSA-6799202 | ACCEPT | Summary: Reactome TAS localization to the mitochondrial inner membrane, from a "Complex I biogenesis" assembly step. Consistent with the experimentally supported inner membrane location. Accept. Reason: A further assembly step of the same inner-membrane pathway (the 315 kDa and 370 kDa subcomplexes associating). Consistent with the experimental localization; accepted as corroboration. Supporting Evidence: Reactome:R-HSA-6799202 The 315kDa and 370kDa subcomplexes associate to form a 550kDa complex |
| GO:0005743 mitochondrial inner membrane | TAS Reactome:R-HSA-6799203 | ACCEPT | Summary: Reactome TAS localization to the mitochondrial inner membrane, from the "Complex I biogenesis" step in which the IP subcomplex binds NDUFAF3, NDUFAF4 and TIMMDC1 to form Intermediate 1. This is the assembly step in which NDUFAF3 directly participates. Consistent with the inner membrane location. Accept. Reason: This is the step in which NDUFAF3 itself acts: the IP subcomplex, centred on the core iron-sulfur subunits NDUFS2 and NDUFS3, binds NDUFAF3, NDUFAF4 and TIMMDC1 to form Intermediate 1. It matches the UniProt SUBUNIT record that NDUFAF3 interacts with NDUFAF4, NDUFS2 and NDUFS3, and is the most directly relevant of the Reactome rows. Accepted. Supporting Evidence: Reactome:R-HSA-6799203 around the core subunits NADH dehydrogenase [ubiquinone] iron-sulfur proteins 2 file:human/NDUFAF3/NDUFAF3-uniprot.txt Interacts with NDUFAF4, NDUFS2 and NDUFS3. |
| GO:0005634 nucleus | IDA GO_REF:0000054 | KEEP AS NON CORE | Summary: IDA nuclear localization from a LIFEdb high-throughput GFP-fusion localization screen (GO_REF:0000054). Overexpressed fusion-protein screens are prone to mislocalization artifacts, and the biologically established location of NDUFAF3 is the mitochondrial inner membrane, where it acts as a complex I assembly factor. Per policy this experimental annotation is not removed. The nuclear pool has no established function and likely reflects the screening method, so it is retained as non-core (matching the IEA nucleus annotation), not treated as a core/functional location. Reason: LIFEdb localizes N- or C-terminally GFP-tagged, overexpressed constructs; for a nucleus-encoded mitochondrial protein whose targeting information lies at the N-terminus, a tag placed there is a well-known source of mislocalization, and overexpression saturates import. No nuclear function has ever been reported for NDUFAF3, and the only physiological location supported by a directed assay is the inner membrane. Per policy an experimental annotation is not removed, so it is kept as non-core. No primary publication is attached to this GO_REF, so no literature quote is offered. Supporting Evidence: file:human/NDUFAF3/NDUFAF3-uniprot.txt Essential factor for the assembly of mitochondrial |
| GO:0005515 protein binding | IPI PMID:19463981 Mutations in NDUFAF3 (C3ORF60), encoding an NDUFAF4 (C6ORF66... | REMOVE | Summary: IPI protein binding from the primary NDUFAF3 disease/assembly study, capturing the physiologically important interactions of NDUFAF3 with NDUFAF4 (Q9P032), NDUFS2 (O75306) and NDUFS3 (O75489). The generic GO:0005515 term is uninformative on its own, but these interactions are the mechanistic basis of NDUFAF3's assembly-factor role. Keep the interactions (captured in core_functions) but remove the bare term. Reason: Unlike the interactome-derived rows, these interactions were established in a hypothesis-driven study and are the mechanistic substance of NDUFAF3's function: NDUFAF4 is its obligate partner factor and NDUFS2/NDUFS3 are the iron-sulfur core subunits around which the first assembly intermediate nucleates. The evidence is strong; it is the term that is inadequate. Removing GO:0005515 does not dispute these interactions - they are carried in core_functions - and the interaction data alone do not license a specific molecular function term for NDUFAF3, which has no catalytic activity of its own. Supporting Evidence: PMID:19463981 Furthermore, we show that NDUFAF3 tightly interacts with NDUFAF4 (C6ORF66), a protein previously implicated in complex I deficiency. |
| GO:0005743 mitochondrial inner membrane | IDA PMID:19463981 Mutations in NDUFAF3 (C3ORF60), encoding an NDUFAF4 (C6ORF66... | ACCEPT | Summary: Direct experimental (IDA) localization of NDUFAF3 to the mitochondrial inner membrane, from the primary study. This is the authoritative localization and the compartment where NDUFAF3 functions as a complex I assembly factor. Accept as core. Reason: The primary study localized NDUFAF3 directly, and UniProt records the inner membrane as the one experimentally evidenced compartment for this protein (ECO:0000269|PubMed:19463981), in contrast to the by-similarity nucleus entry. This is the compartment in which the assembly reactions occur, so it is the core localization. Supporting Evidence: file:human/NDUFAF3/NDUFAF3-uniprot.txt Nucleus {ECO:0000250}. Mitochondrion inner membrane {ECO:0000269|PubMed:19463981}. PMID:19463981 We found that NDUFAF3 is a genuine mitochondrial complex I assembly |
| GO:0032981 mitochondrial respiratory chain complex I assembly | IMP PMID:19463981 Mutations in NDUFAF3 (C3ORF60), encoding an NDUFAF4 (C6ORF66... | ACCEPT | Summary: Direct experimental (IMP) evidence for the core biological process: NDUFAF3 loss-of-function in patients causes isolated complex I deficiency, and GFP-complementation demonstrated that NDUFAF3 is required for complex I assembly. This is the central, best-supported function of the gene. Accept as core. Reason: The strongest evidence in the record: five complex I-deficient patients from three families all carried NDUFAF3 mutations, and pathogenicity was established prospectively by NDUFAF3-GFP baculovirus complementation of patient fibroblasts, so the phenotype is attributable to loss of this protein rather than to linkage. Unlike a substrate, NDUFAF3 performs part of the work of the process - it binds and stabilizes the nascent IP subcomplex - which is what licenses `involved_in`. Supporting Evidence: PMID:19463981 We found that NDUFAF3 is a genuine mitochondrial complex I assembly protein that interacts with complex I subunits. |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: What is the precise molecular activity of NDUFAF3 during complex I assembly (e.g. scaffolding/bridging of NDUFS2-NDUFS3 with NDUFAF4, membrane insertion chaperoning), and could a specific molecular function term beyond generic protein binding be defined for it?
Suggested experts: Nijtmans LG, Saada A
Q: NDUFAF3 is homologous to the bacterial SecF/SecD/YajC membrane-insertion gene cluster. Does NDUFAF3 retain any insertase or insertase-adaptor activity, i.e. does it help thread the mitochondrially encoded ND subunits into the inner membrane, or is the homology purely a shared fold with no retained mechanism?
Suggested experts: Huynen MA, Nijtmans LG
Q: UniProt records a nuclear localization for NDUFAF3 by similarity only, and a LIFEdb GFP-fusion screen produced an IDA nucleus annotation. Is there any endogenous nuclear pool of NDUFAF3 with a function, or should both annotations be treated as artefacts of overexpressed tagged constructs?
Q: The two NDUFAF3 missense variants G77R and R122P and the Leigh-syndrome variant A165V give very different clinical severities. Do they differ in which partner interaction they disrupt - NDUFAF4 binding versus NDUFS2/NDUFS3 binding - and does that explain the phenotypic range from fatal neonatal disease to Leigh syndrome?
Suggested experts: Saada A
Q: Splice isoform b (Q9BU61-2) lacks residues 1-57, the region that would normally carry mitochondrial targeting information, yet it still binds NDUFAF4 in IntAct (NbExp=8) - so the 1-57 region is not required for the partner-factor interaction. Is this isoform translated in vivo, where does it localize, and does it have any assembly-factor activity, or does it bind NDUFAF4 outside the mitochondrion?
Experiment: Structural and biochemical characterization (e.g. cryo-EM of stalled assembly intermediates, crosslinking-MS) of the NDUFAF3-NDUFAF4-NDUFS2-NDUFS3 module, combined with structure-guided separation-of-function mutants tested for complex I assembly rescue in NDUFAF3-null cells.
Hypothesis: NDUFAF3 bridges the NDUFAF4 partner factor to the NDUFS2/NDUFS3 iron-sulfur core subunits to nucleate the early membrane-arm assembly intermediate.
Type: structural and functional interaction analysis
Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)