function_assignment)Hypothesis under evaluation: Human GCDH (Q92947) participates in fatty acid oxidation (GO:0019395) and specifically fatty acid beta-oxidation using acyl-CoA dehydrogenase (GO:0033539).
Human GCDH (glutaryl-CoA dehydrogenase, UniProt Q92947) should not carry the fatty-acid-oxidation biological-process terms GO:0019395 or GO:0033539 as descriptions of its physiological function. The seed hypothesis raises several legitimate nuances — that GCDH is chemically a member of the acyl-CoA dehydrogenase (ACAD) superfamily, that glutaryl-CoA is a dicarboxylic acyl-CoA that is not automatically excluded from fatty-acid chemistry, and that GCDH can dehydrogenate a medium-chain fatty acyl-CoA in vitro. Each of those sub-claims is true at the molecular/chemistry level. But none of them rescues the disputed biological-process annotations, because the specific evidence chain behind those annotations collapses on inspection.
The single experimental (IDA) anchor for GO:0033539 is a misattribution: it cites PMID:25416781(https://pubmed.ncbi.nlm.nih.gov/25416781/), a paper about the mitochondrial lysine methyltransferase METTL20 and its methylation of ETFβ, in which GCDH appears only as an electron donor — not as the subject of a β-oxidation assay. The accompanying phylogenetic (IBA) annotation is circular, anchored (withFrom) on GCDH's own misattributed IDA. The second term, GO:0019395, is an IEA ortholog-transfer artifact from rat that is conspicuously absent from the genuine fatty-acid-oxidation enzymes. Meanwhile GCDH's real reaction — a single oxidative decarboxylation of the dicarboxylic glutaryl-CoA to crotonyl-CoA + CO₂ within lysine/tryptophan catabolism — is categorically distinct from the iterative β-oxidation spiral that GO:0033539 defines.
Verdict: over-annotated / refuted for the process terms, with a narrow, qualified caveat that GCDH does possess a low-level in vitro fatty acyl-CoA dehydrogenation capacity (e.g., on hexanoyl-CoA) that belongs, if anywhere, at the molecular-function level as a non-core promiscuous activity. The recommended curation action is to remove GO:0033539 and GO:0019395 as GCDH functions while retaining glutaryl-CoA dehydrogenase activity (GO:0004361) and lysine/tryptophan catabolism as core.
The QuickGO annotation set for Q92947 lists GO:0033539 (fatty acid beta-oxidation using acyl-CoA dehydrogenase) with evidence code IDA, referenced to PMID:25416781(https://pubmed.ncbi.nlm.nih.gov/25416781/) (Malecki et al., J Biol Chem 2015), supplemented by an IBA from GO_REF:0000033. On retrieval, PMID:25416781 is a characterization of human METTL20 as a mitochondrial lysine methyltransferase that methylates ETFβ (the beta subunit of electron transfer flavoprotein). GCDH is mentioned only as one of the dehydrogenases whose ability to transfer electrons to ETF is diminished when ETFβ is methylated:
"METTL20-mediated methylation of ETFβ in vitro reduced its ability to receive electrons from the medium chain acyl-CoA dehydrogenase and the glutaryl-CoA dehydrogenase."
That sentence is the entire GCDH-relevant content of the cited reference. There is no assay of GCDH performing iterative fatty-acid β-oxidation anywhere in the paper. The seed hypothesis explicitly asked us to recover the full assay methods/results and to determine "which substrate was tested with GCDH versus MCAD." The answer is that neither GCDH nor MCAD was assayed for β-oxidation in that study; both were used as electron donors to test the functional consequence of ETFβ methylation. The IDA is therefore a text-derived misattribution — most plausibly generated by co-occurrence of "glutaryl-CoA dehydrogenase" with an acyl-CoA-dehydrogenase/ETF context, rather than by any demonstrated β-oxidation experiment. Notably, UniProt appears to have attached the same PMID:25416781-derived GO:0033539 to ACADM (P11310) as well, consistent with a bulk, text-derived assignment rather than curated biology.
The GO definition of GO:0033539 specifies a fatty acid β-oxidation pathway that "begins with the addition of coenzyme A to a fatty acid, and ends when only two or three carbons remain (as acetyl-CoA or propionyl-CoA)." This is an iterative, multi-enzyme spiral that shortens a fatty acyl chain two carbons at a time (dehydrogenation → hydration → dehydrogenation → thiolysis, repeated).
GCDH's actual catalytic reaction (GO:0004361, RHEA:13389) is fundamentally different:
glutaryl-CoA + oxidized ETF ──► (2E)-butenoyl-CoA (crotonyl-CoA) + CO2 + reduced ETF
This is a single dehydrogenation coupled to a decarboxylation acting on a dicarboxylic CoA thioester. The physiological substrate, glutaryl-CoA, is generated in the catabolism of L-lysine, hydroxylysine, and L-tryptophan — not from fatty-acid activation. Multiple primary sources confirm this:
GCDH does not run iterative two-carbon-removal cycles yielding acetyl-/propionyl-CoA from a fatty acid; it performs one turnover event on one dicarboxylic substrate. The crotonyl-CoA it produces is a downstream metabolic convergence point (crotonyl-CoA is also a β-oxidation intermediate), but producing a shared intermediate is not the same as running the β-oxidation process. This is precisely the "demonstrated capability versus only downstream metabolic convergence" distinction the seed hypothesis asked us to make — and it resolves against process participation.
Consistent with GCDH being a bona fide member of the ACAD superfamily, it retains detectable activity on straight-chain fatty acyl-CoAs in vitro:
The critical curation point: this establishes a molecular capability (acyl-CoA dehydrogenase activity on a fatty substrate), used here largely as a tool to dissect the enzyme's mechanism, at low relative activity compared with its glutaryl-CoA turnover. It does not establish that GCDH participates in the fatty-acid β-oxidation process in vivo. The seed hypothesis rightly warns against equating "low relative activity" with "absence of capacity" — and that warning is respected here: the capacity is acknowledged and belongs at the MF level as a non-core/promiscuous activity, but it does not license a BP process annotation.
Inspection of the QuickGO withFrom provenance for Q92947 shows the GO:0033539 IBA (GO_Central, GO_REF:0000033) has withFrom = ['PTN005130278', 'Q92947']. In other words, the phylogenetically inferred annotation is anchored on GCDH's own experimental annotation — the very IDA shown in F001 to be a misattribution from PMID:25416781. This is circular provenance: the IBA does not add independent evidence; it re-propagates the erroneous IDA.
Two further observations reinforce the over-annotation pattern:
Importantly, the phylogenetic machinery does correctly identify GCDH's true function: the GCDH-clade PANTHER node PTN005130278 carries the GCDH-specific molecular-function IBA GO:0004361 (glutaryl-CoA dehydrogenase activity), confirming it is the glutaryl-CoA-dehydrogenase ortholog node (distinct from ACADM's node PTN002535634). So the correct MF propagates cleanly along the tree, while the β-oxidation BP term rides along as leakage anchored on a bad IDA.
The evidence produces a clean three-way separation between what GCDH actually does, what it is chemically capable of in a test tube, and what has been erroneously annotated.
GCDH (Q92947) — what the evidence supports
─────────────────────────────────────────────────────────────────────────────
PHYSIOLOGICAL (core) IN VITRO CAPACITY (non-core)
────────────────────── ──────────────────────────────
L-lysine / L-tryptophan catabolism promiscuous ACAD activity
│ on straight-chain acyl-CoA
▼ (e.g. hexanoyl-CoA, C6)
glutaryl-CoA (dicarboxylic) │
│ GCDH: single oxidative ▼
│ decarboxylation (RHEA:13389) simple dehydrogenation
▼ ETF = electron acceptor (2,3-enoyl-CoA), no
crotonyl-CoA + CO2 decarboxylation, low rate
│
▼ (downstream convergence only)
crotonyl-CoA also appears in β-oxidation
→ NOT evidence GCDH runs β-oxidation
─────────────────────────────────────────────────────────────────────────────
ERRONEOUS ANNOTATIONS (should be removed / not asserted as GCDH function)
──────────────────────────────────────────────────────────────────────────
GO:0033539 IDA ← PMID:25416781 = METTL20/ETFβ paper (misattribution)
GO:0033539 IBA ← withFrom Q92947 itself = circular on the bad IDA
GO:0019395 IEA ← Ensembl ortholog transfer from rat = transfer artifact
Interpretation. GCDH is definitively a glutaryl-CoA dehydrogenase (MF GO:0004361) that catalyzes one oxidative decarboxylation step within lysine/tryptophan degradation (BP GO:0019477 and related). Its membership in the ACAD superfamily gives it a residual ability to dehydrogenate short/medium-chain fatty acyl-CoAs in vitro — a fact exploited by mechanistic studies — but this is a low-level side activity, not a physiological role in energy-yielding fatty-acid β-oxidation. The two fatty-acid process annotations trace to (a) a single misattributed IDA from an unrelated methyltransferase paper, (b) an IBA that circularly re-uses that IDA, and (c) an IEA ortholog transfer. None reflects a demonstrated β-oxidation function.
The seed hypothesis's most defensible sub-claims — that dicarboxylate chemistry alone doesn't exclude fatty-acid oxidation (ChEBI places glutaryl-CoA(5-) under omega-carboxy-(fatty acyl)-CoA(5-)), and that GCDH can turn over a fatty acyl-CoA — are true at the chemistry/molecular level. But they do not rescue the process terms, because GO:0033539 has a specific iterative-spiral definition that GCDH does not satisfy, and GO:0019395 is present only via an automated artifact.
| Citation | Evidence type | Supports / Refutes / Qualifies | Claim tested | Key finding | Context | Confidence & limitations |
|---|---|---|---|---|---|---|
| PMID:25416781(https://pubmed.ncbi.nlm.nih.gov/25416781/) | Review of cited IDA source (database provenance) | Refutes (as β-oxidation evidence) | Does the GO:0033539 IDA reference actually assay GCDH β-oxidation? | Paper characterizes METTL20 methylation of ETFβ; GCDH only mentioned as an electron donor whose ETF transfer drops with methylation. No β-oxidation assay of GCDH. | Human, in vitro (methyltransferase focus) | High that it is a misattribution; abstract/discussion-level access, but GCDH content is confined to the ETFβ electron-transfer sentence |
| GO/QuickGO definition of GO:0033539; RHEA:13389 | Ontology definition + reaction chemistry | Refutes | Does GCDH's reaction match the iterative β-oxidation definition? | GO:0033539 = iterative CoA-fatty-acid spiral ending in acetyl/propionyl-CoA; GCDH does a single decarboxylating dehydrogenation of a dicarboxylate. | Definitional / reaction database | High; definitional comparison |
| PMID:15274622(https://pubmed.ncbi.nlm.nih.gov/15274622/) | Structural (crystal structures) | Refutes / qualifies | Nature of GCDH reaction | GCDH decarboxylates glutaryl-CoA to CO₂ + crotonyl-CoA; distinct from β-oxidation. | Human GCDH protein | High; direct structural/mechanistic |
| PMID:37198486(https://pubmed.ncbi.nlm.nih.gov/37198486/) | Cellular / pathway | Refutes (redirects to lysine catabolism) | Physiological process of GCDH | GCDH is the crotonyl-CoA-producing enzyme of lysine catabolism in human cells. | Human glioblastoma cells | High; confirms amino-acid catabolism context |
| PMID:17176108(https://pubmed.ncbi.nlm.nih.gov/17176108/) | Direct assay (kinetics) | Qualifies | Reaction/kinetics of GCDH | Oxidative decarboxylation of glutaryl-CoA; crotonyl-CoA release is rate-limiting. | Purified enzyme | High; kinetic mechanism |
| PMID:21974953(https://pubmed.ncbi.nlm.nih.gov/21974953/) | Direct assay (substrate specificity) | Qualifies / partially supports (MF) | Is GCDH mechanistically an ACAD with fatty acyl capacity? | Catalytic properties very similar to short/medium-chain ACAD except added decarboxylation. | Rat GCDH, in vitro | High for mechanism; species = rat; relative activity not equated to physiology |
| PMID:11024031(https://pubmed.ncbi.nlm.nih.gov/11024031/) | Direct assay (mutagenesis + kinetics) | Partially supports (MF only) | Can human GCDH dehydrogenate a fatty acyl-CoA? | Hexanoyl-CoA (C6) is a measurable non-decarboxylated alternative substrate. | Human GCDH, in vitro | High for in vitro capacity; used as mechanistic tool, low physiological relevance |
| QuickGO withFrom provenance (Q92947) | Computational/database provenance | Refutes (circularity) | Is the GO:0033539 IBA independent? | IBA withFrom = ['PTN005130278','Q92947'] — anchored on GCDH's own (misattributed) IDA. | GO_Central phylogenetic annotation | High; provenance directly inspected |
| GO annotation distribution (GCDH, IVD, ACADM/S/VL/L) | Computational/database | Refutes (paralog leakage) | Is GO:0033539 specific to true β-ox enzymes? | Term over-propagated to amino-acid-catabolism ACADs (GCDH, IVD) alongside true β-ox ACADs. | Cross-gene comparison | Medium-high; pattern-level evidence |
How the literature bears on the findings: The two crystallographic/kinetic papers (PMID:15274622(https://pubmed.ncbi.nlm.nih.gov/15274622/), PMID:17176108(https://pubmed.ncbi.nlm.nih.gov/17176108/)) and the human cellular study (PMID:37198486(https://pubmed.ncbi.nlm.nih.gov/37198486/)) collectively pin GCDH's physiological role to a single decarboxylating dehydrogenation within lysine catabolism, challenging the β-oxidation process claim. The two substrate-specificity/mutagenesis papers (PMID:21974953(https://pubmed.ncbi.nlm.nih.gov/21974953/), PMID:11024031(https://pubmed.ncbi.nlm.nih.gov/11024031/)) support only a molecular-level acyl-CoA-dehydrogenase capacity, not process participation. The cited IDA source (PMID:25416781(https://pubmed.ncbi.nlm.nih.gov/25416781/)) is the linchpin misattribution that, together with the circular IBA and the IEA transfer, generated the disputed annotations.
Lead (requires curator verification):
| GO term | Aspect | Current evidence on Q92947 | Recommended action |
|---|---|---|---|
| GO:0033539 — fatty acid beta-oxidation using acyl-CoA dehydrogenase | BP | IDA (PMID:25416781, misattributed) + IBA (circular, withFrom Q92947) | Remove as a GCDH process annotation. The IDA reference does not support β-oxidation; the IBA re-uses that IDA. |
| GO:0019395 — fatty acid oxidation | BP | IEA (Ensembl ortholog transfer from rat) | Remove / do not assert. Ortholog-transfer artifact; absent from the genuine FA-oxidation enzymes. |
| GO:0004361 — glutaryl-CoA dehydrogenase activity | MF | IDA + IBA on node PTN005130278 | Retain — this is GCDH's core molecular function. |
| GO:0019477 / GO:0006554 — L-lysine catabolic process (glutaryl-CoA / lysine degradation) | BP | Supported by cellular evidence (PMID:37198486) and enzyme role | Retain / strengthen — this is GCDH's core biological process. |
| Generic acyl-CoA dehydrogenase activity (e.g., GO:0003995) | MF | In vitro hexanoyl-CoA activity (PMID:11024031, PMID:21974953) | Optional, non-core. Could be captured at MF level with a clear "in vitro / promiscuous" note; do not upgrade to a fatty-acid β-oxidation BP term. |
Summary of the curation call: The evidence supports retaining GCDH's molecular function (glutaryl-CoA dehydrogenase) and its lysine/tryptophan-catabolism process and removing the two fatty-acid-oxidation biological-process terms (GO:0033539, GO:0019395). If a curator wishes to preserve the demonstrated in vitro fatty acyl-CoA activity, it belongs as a non-core MF-level note, never as a β-oxidation process annotation. Note that this recommendation targets the biological-process claim specifically; it does not dispute GCDH's identity as an ACAD-fold enzyme.
The immediate molecular function tested is flavin-dependent oxidative decarboxylation of glutaryl-CoA (a dicarboxylic CoA thioester) to crotonyl-CoA + CO₂, with ETF as the electron acceptor — an ACAD-type dehydrogenation plus a decarboxylation. This is a discrete, single catalytic event.
Separating direct activity from downstream/inferred effects:
All items below are leads requiring curator verification.
Candidate action changes:
- Remove GO:0033539 (fatty acid beta-oxidation using acyl-CoA dehydrogenase) IDA/IBA from Q92947 — IDA misattributed to PMID:25416781; IBA circular.
- Remove / do-not-assert GO:0019395 (fatty acid oxidation) IEA — ortholog-transfer artifact.
- Retain GO:0004361 (glutaryl-CoA dehydrogenase activity) as core MF.
- Retain / strengthen L-lysine catabolic process (GO:0019477 and related) as core BP.
- Optional non-core MF note: generic acyl-CoA dehydrogenase activity, flagged as in vitro/promiscuous only.
Candidate references with snippets to verify:
- PMID:25416781 — "METTL20-mediated methylation of ETFβ in vitro reduced its ability to receive electrons from the medium chain acyl-CoA dehydrogenase and the glutaryl-CoA dehydrogenase" → shows GCDH is only an electron donor; not a β-oxidation assay.
- PMID:15274622 — "GCD catalyzes decarboxylation of glutaryl-CoA to produce CO(2) and crotonyl-CoA" → confirms single decarboxylating dehydrogenation, not β-oxidation.
- PMID:37198486 — "crotonyl-coenzyme A (crotonyl-CoA)-producing enzyme glutaryl-CoA dehydrogenase (GCDH)" → amino-acid (lysine) catabolism context.
- PMID:21974953 — "catalytic properties very similar to those of short-chain and medium-chain acyl-CoA dehydrogenase except its additional decarboxylation reaction" → mechanistic ACAD, supports MF-level acyl-CoA capacity only.
- PMID:11024031 — "alternative substrates, hexanoyl-CoA and glutaramyl-CoA, which are not decarboxylated" → in vitro C6 fatty acyl-CoA activity, non-core.
Suggested curator questions:
- Does the full text of PMID:25416781 contain any GCDH β-oxidation assay? (Expected: no.)
- Is GO:0033539 on IVD (via PMID:3597357) similarly a co-purification carry-over?
Suggested experiments: GCDH-perturbation fatty-acid-oxidation flux assay; quantitative kinetics comparing glutaryl-CoA and fatty acyl-CoA substrates.
The seed hypothesis raises legitimate nuances — GCDH is chemically an ACAD, glutaryl-CoA is a dicarboxylic acyl-CoA that ChEBI classifies under omega-carboxy-(fatty acyl)-CoA, and the enzyme can dehydrogenate a medium-chain fatty acyl-CoA in vitro — and those nuances correctly caution against dismissing all fatty-acyl capacity out of hand. But they do not sustain the disputed process annotations. The GO:0033539 IDA is misattributed to an unrelated METTL20/ETFβ paper; its IBA is circularly anchored on that same bad IDA; GO:0019395 is an IEA ortholog-transfer artifact; and GCDH's actual chemistry is a single oxidative decarboxylation within lysine/tryptophan catabolism, not the iterative β-oxidation spiral GO:0033539 defines. Recommendation: remove GO:0033539 and GO:0019395 as GCDH functions; retain glutaryl-CoA dehydrogenase activity and lysine catabolism as core; and, at most, represent any residual fatty-acyl activity as a non-core in vitro molecular function.