liuC (Q9I298, PA2013) notes

Review session 2026-10-28 (PseudoCAP IBA-seed protocol, worklist rank 40)

No deep-research provider was available. The two PseudoCAP-cited papers (PMID:16517656, PMID:16820476)
are abstract-only and the PubMed MCP returned no PMC full text. I cached four further papers on the Liu/Atu
pathways with just fetch-pmid (PMID:16272386, PMID:18625020, PMID:19459965, PMID:27271456); all are
abstract-only.

Identity

P. aeruginosa PAO1 locus PA2013, UniProt Q9I298 (unreviewed), 265 aa; submitted name
"3-methylglutaconyl-CoA hydratase" (ECO:0000313, not evidence).
Enoyl-CoA hydratase/isomerase (crotonase) family: Pfam PF00378 (ECH_1), PROSITE PS00166
(ENOYL_COA_HYDRATASE), InterPro IPR001753, IPR018376; PANTHER PTHR42964:SF1.
Third gene of the liuRABCDE cluster (PA2011-PA2016); LiuB/LiuD (PA2014/PA2012) are the
methylcrotonyl-CoA carboxylase subunits PMID:16272386, LiuE (PA2011) the HMG-CoA lyase PMID:19459965.

Pathway logic

Leucine/isovalerate catabolism: isovaleryl-CoA (LiuA) -> 3-methylcrotonyl-CoA -> carboxylation (LiuB/LiuD)
-> 3-methylglutaconyl-CoA -> hydration (LiuC, the step this gene is named for) -> HMG-CoA -> cleavage
(LiuE) -> acetyl-CoA + acetoacetate. Acyclic monoterpenoids (citronellol, geraniol) are degraded by the Atu
pathway, which joins the Liu pathway at methylcrotonyl-CoA PMID:16272386. LiuC therefore acts in both pathways in
P. aeruginosa, and the genetics agree: PMID:18625020; PMID:16517656.

The terpene role is not a property of the hydratase: LiuC acts on 3-methylglutaconyl-CoA, which is a
leucine catabolic intermediate. Terpenoid catabolism uses LiuC only where an Atu pathway feeds
methylcrotonyl-CoA into the Liu pathway. Most organisms with a LiuC-type hydratase have leucine catabolism
and no Atu pathway; P. putida KT2440 grows on leucine but not on acyclic terpenes PMID:16820476.

Term precision

GO:0046247 terpene catabolic process is defined for terpene hydrocarbons. Citronellol and geraniol are
monoterpenoid alcohols (C10), so GO:0016100 monoterpenoid catabolic process (is_a terpenoid catabolic
process is_a isoprenoid catabolic process GO:0008300) is the accurate term.

Molecular function evidence

No purified-enzyme assay of P. aeruginosa LiuC was found. The PseudoCAP MF row is IMP to PMID:16517656, whose
abstract does not mention a liuC mutant or a hydratase assay. The MF is nevertheless well supported by
pathway logic (it is the only hydratase in the liu cluster, whose carboxylase product is
3-methylglutaconyl-CoA and whose lyase substrate is HMG-CoA), by the crotonase-family signature, and by the
characterised Myxococcus LiuC homolog PMID:27271456 (a different paralog, which that paper notes is
more similar to human MGCH than to bacterial forms).

Annotation decisions

GO:0004490 methylglutaconyl-CoA hydratase activity (IMP PseudoCAP; IBA): ACCEPT, core MF (deferring on
the IMP full text).
GO:0006552 L-leucine catabolic process (IMP): ACCEPT, core BP.
GO:0046247 terpene catabolic process (IMP PseudoCAP; IBA): MODIFY to GO:0016100 monoterpenoid catabolic
process; KEEP the claim for PAO1, but the IBA should not have been propagated (GO_CENTRAL defect).
GO:0008300 isoprenoid catabolic process (IMP): MODIFY to GO:0016100.
GO:0003824 catalytic activity (IEA InterPro): MODIFY to GO:0004490.

PseudoCAP IBA-seed assessment

QuickGO (2026-10-28, withFrom=PANTHER:PTN001733227) returns 13 leaves, each with GO:0004490 and GO:0046247:
PAO1 LiuC (Q9I298) and AtuE (Q9HZV7, isohexenylglutaconyl-CoA hydratase of the Atu pathway), B. subtilis
PksH (P40805, bacillaene synthase dehydratase, which per UniProt (ECO:0000269) dehydrates an HMG group
tethered to the carrier protein PksL, not a CoA thioester), Coxiella PksH-like and CBU_0976, M.
tuberculosis EchA7, Shewanella LiuC, Xanthomonas FadB and XCC1833, Bradyrhizobium bll0116, Streptomyces
SCO4384, Chloroflexus Caur_1565 and the choanoflagellate Monosiga brevicollis.

PTN node	node term	n_seeds	leaf IBAs	seed annotation (term, evidence, PMID)	verdict	rationale
PTN001733227	GO:0046247 terpene catabolic process	1 (sole)	13 (10 taxa)	GO:0046247, IMP, PMID:16820476 (PseudoCAP)	OVER-SPECIFIC	For PAO1 the claim holds (liu mutants cannot use acyclic terpenes, PMID:18625020; the Atu pathway enters the Liu pathway at methylcrotonyl-CoA, PMID:16272386), though GO:0016100 monoterpenoid catabolic process is the accurate term. But terpenoid catabolism is an organism-level pathway context (an Atu pathway feeding the leucine pathway), not a property of the hydratase. It is not inherited by the clade: Mycobacterium, Bacillus PksH, Coxiella, Xanthomonas, Streptomyces, Chloroflexus and Monosiga have no Atu pathway, and even P. putida, with a Liu pathway, cannot grow on acyclic terpenes. The IBD should be withdrawn.
PTN001733227	GO:0004490 methylglutaconyl-CoA hydratase activity	1 (sole)	13 (10 taxa)	GO:0004490, IMP, PMID:16517656 (PseudoCAP)	UNVERIFIABLE	The seed is a mutant-based MF inference from an abstract-only paper that does not mention liuC or a hydratase assay. The activity is very likely right for LiuC by pathway logic. The node term is too specific for the clade: it includes AtuE (a different substrate, isohexenylglutaconyl-CoA) and B. subtilis PksH, an experimentally characterised polyketide-pathway dehydratase acting on a carrier-protein-bound HMG group rather than on HMG-CoA. A generic hydro-lyase/enoyl-CoA hydratase term, or a node restricted to LiuC orthologs, would be safer.

PANTHER tree check (2026-11-05)

Topology of PTN001733227 fetched with projects/PSEUDOCAP/panther_tree.py node PTN001733227 --depth 4 / --leaves / --contains (PANTHER treeinfo for PTHR42964).

PTN001733227: SPECIATION, placed at LUCA, 13 leaves, all PTHR42964:SF1.

child node	event / taxon	leaves	subfamilies	content	verdict
PTN000941888	LEAF Monosiga brevicollis	1	SF1	enoyl-CoA hydratase	no Liu/Atu pathway
PTN000234935	SPECIATION Eubacteria	12	SF1	all bacteria	—
└ PTN008092237	SPECIATION Terrabacteria	4	SF1	B. subtilis PksH (bacillaene dehydratase), Chloroflexus, M. tuberculosis EchA7, S. coelicolor	substrates not 3-methylglutaconyl-CoA
└ PTN000234939	SPECIATION Proteobacteria	8	SF1	Bradyrhizobium bll0116 + PTN001733230	—
└ PTN001733230	SPECIATION Gammaproteobacteria	7	SF1	Coxiella pair PTN001733228, Shewanella LiuC, Xanthomonas pair PTN002873558 (one a p-hydroxycinnamoyl-CoA hydratase/lyase), PAO1 pair PTN002452381	narrowest clade holding all three MG-CoA hydratases; still heterogeneous
└ PTN002452381	DUPLICATION P. aeruginosa	2	SF1	AtuE (PTN002452382) + LiuC (PTN002452383)	the only clade in which acyclic terpenoid catabolism holds

Conclusions:

The node's event placement is LUCA, so both IBDs assert their term as ancestral
to all life. For a substrate-specific crotonase activity, and still more for a
pathway context that depends on an organism having an Atu cluster, that is not
supportable.
LiuC's own sister leaf is AtuE, a hydratase with a different substrate
(--contains Q9I298,Q9HZV7 -> PTN002452381, 2 leaves). The smallest clade holding
the three methylglutaconyl-CoA hydratases of the family (PAO1 Q9I298, Shewanella
Q8EFS1, Coxiella Q83CX5) is PTN001733230, which also holds AtuE, Coxiella PksH
and the two Xanthomonas enzymes. So there is no LiuC-ortholog-only subnode;
GO:0004490 belongs at the leaves, and the node should carry a generic enoyl-CoA
hydratase/hydro-lyase term.
For GO:0046247 the tree does supply one defensible target: PTN002452381, the PAO1
LiuC/AtuE duplication pair, since both participate in acyclic terpene utilisation in
P. aeruginosa. Everything above that is over-propagation.

Both defects raised to CONFIRMED (GO:0004490 SUSPECTED -> CONFIRMED, GO:0046247
PROBABLE -> CONFIRMED). Annotation actions unchanged: ACCEPT on the GO:0004490 IBA
(correct for this gene) and MODIFY to GO:0016100 on the GO:0046247 rows.