Repository navigation
Reaction and GPR corrections (redox, choline, lactate, prostanoids, halides, methionine synthase) #1109
Description
This issue collects smaller reaction and GPR errors: the lactate dehydrogenase family, choline oxidation, D-lactate, hydroxypyruvate reduction, PTGR3, HCl/HBr, the P5C ring opening and a free-cobalamin methionine synthase. Some of these reactions created reducing equivalents or used enzymes in the wrong compartment or with the wrong cofactor. Some also carried the identifiers of a different reaction. Bile acid steps are covered in #1108, and transporters in #1111.
1. Lactate dehydrogenase family (#1106)
LDHD is a mitochondrial, FAD-dependent (R)-2-hydroxyacid dehydrogenase (UniProt Q86WU2; PMID 30931947, PMID 38373542). It does not belong on cytosolic NADH reactions.
| Change | Model entities | Evidence |
|---|---|---|
Glyoxylate oxidation to oxalate: LDHD removed and LDHA added. GPR: LDHA or LDHB or LDHC or LDHAL6A or LDHAL6B (#1106) |
MAR08780 | Recombinant human LDHA oxidises glyoxylate with NAD+ (PMID 16198644). LDHC and the LDHAL6 genes were kept as they were (not tested) |
Hydroxypyruvate reductase. Before: 3 H+ + 2 hydroxypyruvate + NADH -> NAD+ + 2 (S)-glycerate (charge -2). After: 1:1:1 -> 1:1. GPR: 6 LDH genes -> LDHA or LDHB. MetaNetX MNXR158927 -> MNXR100701. The L-glycerate ids MNXR100701/HPYRR2x were moved here from GRHPR's D-glycerate reaction MAR08782 (#1106) |
MAR01453, MAR08782 | The old form produced reducing equivalents from nothing. Human recombinant LDHA and LDHB reduce hydroxypyruvate (PMID 16198644); LDH gives L-glycerate (chicken, PMID 196642; L-glyceric aciduria, PMID 5635456) |
| NAD+ D-lactate dehydrogenase (D-lactate + NAD+ <=> pyruvate + NADH, LDHD only): deprecated (#1106) | MAR08512 | Human LDHD uses FAD. NAD-dependent D-LDH (RHEA:16369) is found only in bacteria and invertebrates (PMID 38373542). D-lactate is still oxidised to pyruvate by MAR03859 (LDHD, cytochrome c) |
2. Choline oxidation (#1106)
| Change | Model entities | Evidence |
|---|---|---|
| Cytosolic choline + NAD+ -> betaine aldehyde (CHDH): deprecated. Also deprecated: the two cytosolic betaine aldehyde dehydrogenations (ALDH7A1, NAD+ and NADP+) and cytosolic betaine aldehyde (#1106) | MAR04696, MAR04697, MAR04698, MAM01394c | CHDH is an FAD enzyme of the inner mitochondrial membrane (UniProt Q8NE62). No human NAD+ choline dehydrogenase exists (RHEA:33051 is bacterial). Without MAR04696, the two ALDH7A1 reactions had no substrate. Betaine synthesis still runs MAR08657 -> MAR08441 -> MAR08440 (ALDH7A1 [m]) -> MAR08890 |
3. PTGR3 15-oxoprostaglandin 13-reductase (#1106, #1104)
| Change | Model entities | Evidence |
|---|---|---|
Before: 13,14-dihydro-15-oxo-PGE2 + NADP+ -> 15-oxo-PGE2 + NADPH. After: 15-oxo-PGE2 + NADPH + H+ -> 13,14-dihydro-15-oxo-PGE2 + NADP+. Duplicated GPR ZADH2 or ZADH2 -> PTGR3 (ZADH2). Placeholder names replaced by "15-oxoprostaglandin 13-reductase (15-oxo-PGE2)" |
MAR20018 [g] (#1106), MAR20019 [n] (#1104) | PTGR3 is a 15-oxoprostaglandin 13-reductase (UniProt Q8N4Q0, RHEA:11912) |
Open point: both reactions are isolated, because neither prostaglandin has any other reaction in g or n. UniProt places PTGR3 in the peroxisome, so the compartment still needs a decision.
4. HCl and HBr (#1106)
| Change | Model entities | Evidence |
|---|---|---|
| HCl replaced by chloride + H+ in the GST dehalogenations and the xenobiotics sink. HCl deprecated into chloride, which records MAM02144c and its HMR id as retired | MAM02144c -> MAM01442c; MAR07029, MAR07092, MAR07093, MAR07097, MAR00035 | HCl is fully dissociated in water. The same was done earlier for hydrogen iodide (merged into iodide) |
| "hydrobromic acid" renamed "bromide". HBr ids replaced by bromide ids (CHEBI:15858, KEGG C01324, PubChem 259, SMILES [Br-]) | MAM02143c | Its formula (Br) and charge (-1) were already those of bromide, so only the name and xrefs were wrong |
5. 1-Pyrroline-5-carboxylate ring opening (#1106)
| Change | Model entities | Evidence |
|---|---|---|
| P5C (and 3-hydroxy-P5C) + H2O <=> glutamate 5-semialdehyde: ALDH4A1 and EC 1.2.1.88 removed. Renamed "... hydration (spontaneous)" and flagged spontaneous | MAR03819, MAR03820, MAR08096 | The ring-chain tautomerism is non-enzymatic (RHEA:28234 has no enzyme). ALDH4A1 catalyses the following NAD+ oxidation, which is modelled separately and keeps its gene |
6. Free-cobalamin methionine synthase (#1104)
| Change | Model entities | Evidence |
|---|---|---|
| cob(II)alamin + SAM -> methylcobalamin + SAH (no gene) deprecated into MAR09550. Methylcobalamin (MAM03322c) deprecated | MAR03138, MAM03322c | Methionine synthase is reactivated by MTRR on the enzyme-bound cofactor, which MAR09550 already models (UniProt Q9UBK8). Free methylcobalamin had no other reaction, so MAR03138 could carry no flux |
7. Blocked ER copy of 4-nitrophenol hydroxylation (#1106)
| Change | Model entities | Evidence |
|---|---|---|
| Deprecated into MAR08034. PNP[r] (MAM02754r) and 4-nitrocatechol[r] (MAM01019r) deprecated | MAR20031 | MAR20031 is the same reaction as MAR08034 (PNP + NADPH + H+ + O2 -> 4-nitrocatechol + NADP+ + H2O, CYP2E1, RHEA:26205, MNXR102281), but in the ER, where PNP and 4-nitrocatechol have no other reaction or transport, so it could carry no flux. MAR08034 in the cytosol is connected to PNP sulfation (MAR07688), uptake and exchange. CYP2E1 is an ER enzyme (UniProt P05181); moving MAR08034 to the ER would need new transport reactions and is not done here |
Open points (not changed)
- MAR03859 (LDHD, D-lactate[m] + 2 ferricytochrome c[m]) is the only remaining D-lactate oxidation. Human LDHD transfers electrons from D-lactate directly to cytochrome c: purified enzyme reduces cytochrome c, and LDHD-dependent D-lactate respiration in permeabilised human cells is antimycin-insensitive (PMID 38413804).
- Prostaglandin G/H synthase: PTGS1/2 are ER-membrane enzymes, but the model has the PG1 steps both in [c] (MAR01390, MAR01392, which lead to PGE1 via PTGES3, MAR01394) and in [r] (MAR01391, MAR01393, blocked because PGH1[r] has no other reaction). One compartment should be chosen, with a PGH1 transport if needed. The charge of the peroxidase steps (MAR01307, MAR01392, MAR01393) is fixed in fix: corrections to curation since v2.0.1 from an evidence review #1105.
- Mitochondrial choline import (MAR08657) has no gene.
Implemented in #1106 and #1104. Related corrections of recent curation are in #1105: the LDH GPRs of MAR11476, MAR11479 and MAR05351 (comment on #528), the ubiquinone reactions MAR00483, MAR03838 and MAR08441 (comment on #1015), MAR08607 (comment on #626) and the LIPT1 reaction MAR20168 (comment on #837).
Activity
edkerk commentedon Oct 5, 2026
Fixed in Human 2.1.0.