Cross-check: our FAO module vs the Reactome mitochondrial β-oxidation pathway
← back to the Fatty Acid β-Oxidation project
Our human gene reviews already ingest and adjudicate Reactome at the
annotation level — every human FAO review carries original_reference_id:
Reactome:R-HSA-… TAS annotations for the individual chain-length-specific
reactions, each reviewed (ACCEPT / KEEP_AS_NON_CORE). This page is the
pathway-level cross-check that was missing: does our cross-species module
(MODULE:fatty_acid_beta_oxidation)
agree with how Reactome models the mitochondrial FAO spiral?
Source: the cached Reactome reaction entries under reactome/R-HSA-77*.md
(the Homo sapiens mitochondrial β-oxidation reaction set; catalysts read from
each reaction's "mediated by the '…' of '…'" statement).
Reactome's reaction inventory, by step and chain length
Reactome models the spiral one chain length at a time, with each turn's four
reactions catalysed as follows:
| Chain (acyl-CoA in) | ① dehydrogenase | ② hydratase | ③ 3-OH-acyl-CoA DH | ④ thiolase |
|---|---|---|---|---|
| C16 palmitoyl | VLCAD (77299) | MTP (77301) | MTP (77303) | MTP (77304) |
| C14 myristoyl | — | MTP (77277) | MTP (77283) | MTP (77271) |
| C12 lauroyl | — | ECHS1 (77256) | HADH (77254) | MTP (77309) |
| C10 decanoyl | ACADM (77345) | ECHS1 (77344) | HADH (77342) | MTP (77340) |
| C8 octanoyl | ACADM (77338) | ECHS1 (77333) | HADH (77331) | MTP (77329) |
| C6 hexanoyl | (77327) | ECHS1 (77325) | HADH (77323) | MTP (77321) |
| C4 butanoyl | ACADS (77319) | ECHS1 (77314) | HADH (77312) | — |
("MTP" = Reactome's "Trifunctional Protein" complex.)
Agreements — our module and Reactome match on steps ①–③
- ① Chain-length-specific dehydrogenases. VLCAD for long (C16), ACADM/MCAD
for medium (C8/C10), ACADS/SCAD for short (C4) — exactly our module's
vlcad/mcad/scadvariant sets. - ② + ③ split at the same S/MC ↔ LC boundary. Reactome routes the hydratase
and 3-hydroxyacyl-CoA dehydrogenase steps through ECHS1 and HADH for
C4–C12, and through MTP for C14–C16 — the same short/medium-vs-long split
our module encodes (ECHS1/HADH as the S/MC enzymes; HADHA/MTP as the LC
enzymes). The boundary (C12 handled by ECHS1/HADH, C14 by MTP) is consistent. - The RHEA-chaining gap is a mapping artefact, independently confirmed. Our
module flags step ①→② as a break because the enoyl-CoA hydratase MF term
GO:0004300maps only to RHEA:20724 (the (3E) variant), not the canonical
(2E) crotonase RHEA:16105. Reactome models the same reaction with the correct
(2E) substrate — e.g. R-HSA-77314 "Crotonoyl-CoA + H2O ⇒ (S)-3-Hydroxybutanoyl-CoA" —
confirming the chemistry is (2E)-enoyl-CoA hydration and the gap lies purely
in theGO:0004300 → RHEAmapping, not the biology.
Divergence — the thiolase step ④
Reactome attributes the thiolase step across the entire modelled spiral
(C6→C16) to the "Trifunctional Protein" complex, and neither ACAA2 nor
ACAT1 appears in any cited reaction. Our reviews resolve step ④ more finely:
- ACAA2 (3-ketoacyl-CoA thiolase,
GO:0003988) — the medium/long-chain
straight-chain thiolase of the matrix. - ACAT1 (acetoacetyl-CoA thiolase / T2) — the short-chain / ketone-body
thiolase (the C4 acetoacetyl-CoA step). - HADHB / MTP — the membrane-bound long-chain thiolase only.
So Reactome models steps ②–③ with the chain-length-appropriate soluble S/MC
enzymes (ECHS1/HADH) but keeps step ④ on MTP even at C6–C10, over-attributing
the medium/short-chain thiolysis to the trifunctional complex and omitting the
dedicated soluble thiolases. This is a granularity difference, not an obvious
error — MTP does carry a thiolase (HADHB) activity — but our per-gene reviews
capture the ACAA2 (M/LC) and ACAT1 (SC/ketone) contributions that this Reactome
pathway does not separate out. Flagged here for curators comparing the two
models.
Scope note
The cached Reactome set begins at C16 palmitoyl-CoA (VLCAD) and does not
model the very-long-chain (C17–C22) entry that VLCAD/ACADVL and ACAD9 cover;
our module and the ACAD9/ACADVL reviews (and the ACAD9 GO:0017099→GO:0004466
chain-length MODIFY) address that end of the range separately.
Bottom line
The module and Reactome agree on the pathway skeleton and on steps ①–③
(including the same S/MC-vs-LC enzyme split and the honestly-documented RHEA
mapping gap). The one substantive difference is step ④: our reviews resolve the
mitochondrial thiolase into ACAA2 (M/LC) and ACAT1 (SC/ketone), whereas the
Reactome pathway routes the whole thiolase step through the MTP complex.