Repository navigation
Transport reaction curation: stoichiometry, GPRs and evidence #1111
Description
This issue documents the peroxisomal pore protein PXMP2 (#1100) and corrections to transport reactions that were already in v2.0.1 (#1105, #1106). This issue records, for each change, the ion coupling, direction and GPR that are now in the model and the evidence behind them. Diffusion transports and the spontaneous flag are covered in #1112. The H2O2 aquaporin GPRs (MAR08408, MAR13002, MAR03946) are described there.
1. PXMP2 on cytosol-peroxisome transports (#1100)
What changes: PXMP2 (ENSG00000176894, UniProt Q9NR77) is added as a gene. It becomes the GPR of 40 gene-less cytosol <-> peroxisome uniports of water-soluble solutes up to about 300 Da. Each reaction cites PMID 19352492.
| Solute class | Reactions |
|---|---|
| Organic acids and keto acids | MAR00056 acetoacetate, MAR01566 oxalate, MAR07705 glycolate, MAR07707 glyoxylate, MAR08977 L-lactate, MAR09679 acetate, MAR04469 hydroxypyruvate, MAR07797 2-oxoglutarate, MAR08980 malate, MAR09680 oxaloacetate, MAR06987 succinate, MAR06986 isocitrate, MAR11845 5-guanidino-2-oxopentanoate, MAR11980 5-amino-2-oxopentanoate, MAR08020 1-piperideine-6-carboxylate |
| Dicarboxylic acids | MAR04443 adipate, MAR06821 suberate, MAR06805 sebacate, MAR04515 dodecanedioate, MAR11531 3-hydroxyadipate, MAR11541 3-hydroxysuberate, MAR11536 3-hydroxysebacate |
| Amino acids | MAR03848 glycine, MAR03850 sarcosine, MAR04468 serine, MAR04790 alanine, MAR07640 D-aspartate, MAR08016 lysine, MAR09179 D-ornithine, MAR09181 D-arginine |
| Purines | MAR04647 xanthine, MAR08754 hypoxanthine, MAR08860 urate |
| Phosphates and phosphorylated metabolites | MAR01626 Pi, MAR05346 PPi, MAR04393 DHAP, MAR05414 sn-glycerol-3-phosphate, MAR03948 isopentenyl diphosphate |
| Polyamines | MAR06928 spermine, MAR06937 spermidine |
Why: Pxmp2 forms a non-selective channel in the peroxisomal membrane that passes solutes up to about 300-600 Da (PMID 19352492, mouse). With a gene, ftINIT can keep these transports wherever PXMP2 is expressed. Not included: gases, water and H2O2, which cross without PXMP2 (PMID 31129117, human 293 cells); H+, Na+ and NH4+; lipophilic compounds; O-propanoylcarnitine; anything above 300 Da. A PXMP2 knockout does not change growth in the full model.
Strength of evidence: the channel activity was measured for mouse Pxmp2, and for human PXMP2 it is assigned by orthology. Pxmp2 knockout mice only partly lose peroxisomal permeability (PMID 19352492), but the single-gene GPR makes all 40 transports depend on PXMP2 in a gene deletion.
Open points (not changed):
- Several pore reactions are one-way although a pore is bidirectional: MAR07705, MAR08016, MAR08754, MAR08860, MAR01566, MAR08977, MAR11531, MAR11536, MAR11541.
- Four are still named "(Diffusion)" (MAR04443, MAR04515, MAR06805, MAR06821).
- MAR05344 (H+ + Pi, c <=> x, no gene) overlaps with MAR01626.
- No reason is recorded for excluding O-propanoylcarnitine.
2. Vesicular secretion of 5-hydroxytryptophan and metanephrine (#1106)
The model has no secretory-vesicle compartment, so vesicular secretion is written as a transport from the cytosol to the extracellular space driven by ATP. Two of these reactions have no support:
| Change | Model entities | Evidence |
|---|---|---|
| GPR SLC18A1 or SLC18A2 removed; the reaction is kept without a gene | MAR07634 5-hydroxytryptophan export | 5-HTP is not a VMAT substrate. UniProt Q05940 and P54219 list monoamines, not amino-acid precursors. The reaction is the only route between 5-HTP[c] and 5-HTP[e], so it is not removed. |
| Deprecated | MAR08922 metanephrine vesicular secretion (no gene) | Most plasma metanephrine comes from COMT metabolism within the cell and is released without exocytosis (PMID 9626157, human pheochromocytoma). O-methylated catecholamines are not VMAT substrates. Metanephrine still crosses the plasma membrane by MAR10480. |
Open point: 5-HTP is probably carried across the plasma membrane by an amino-acid transporter such as LAT1 (SLC7A5). That needs a new reaction and is not done here.
3. Bile salt export from the peroxisome removed (#1105)
| Change | Model entities | Evidence |
|---|---|---|
| Deprecated | MAR01857 taurocholate[x] + 2 ATP[x] => taurocholate[e]; MAR01861 taurochenodeoxycholate[x] + 2 ATP[x] => taurochenodeoxycholate[e]; both without a gene | These reactions skipped the cytosol and used peroxisomal ATP. The canalicular bile salt pump ABCB11 uses 1 ATP per bile salt (UniProt O95342, RHEA:50052 taurocholate, RHEA:50064 taurochenodeoxycholate). Export from the cytosol with 1 ATP already exists as MAR01870 and MAR01872 (ABCB11 and others). |
Open point: MAR01702 (25(R)DHCA-CoA r => x, 2 ATP, no gene) is kept, because it is the only way for DHCA-CoA made in the ER (SLC27A5, MAR01699) to reach the peroxisome. The THCA-CoA reaction MAR02285 uses 1 ATP. How bile acyl-CoAs enter the peroxisome needs a separate decision.
4. Ammonia channels (#1105)
| Change | Model entities | Evidence |
|---|---|---|
| GPR RHAG -> RHAG or RHBG or RHCG; reference added; NH4+ Rhea id RHEA:28747 removed (it describes NH4+, not NH3) | MAR04873 NH3 c <=> e | Human RhAG, RhBG and RhCG expressed in oocytes all carry NH3 (PMID 24077989). UniProt Q9H310 (RHBG, basolateral) and Q9UBD6 (RHCG, apical) are reviewed ammonium transporters. RHAG alone is erythroid, so kidney and liver cells had no gene for this reaction. Both genes were already in the model. |
Mitochondrial NH3 (MAR04874) stays diffusion without AQP8; see the comment on #1101.
5. Acetone is not an MCT substrate (#1105)
| Change | Model entities | Evidence |
|---|---|---|
| Deprecated | MAR08365 acetone[c] + H+[i] => acetone[m] + H+[m], GPR SLC16A1 | MCT1 carries monocarboxylates (lactate, pyruvate, ketone body acids) with H+ (UniProt P53985). Acetone is neutral and not a carboxylate. |
| H+ removed, made reversible, flagged as diffusion | MAR08364 acetone[e] <=> acetone[c]; MAR02222 acetone[m] <=> acetone[c] (before: m => e directly) | Small neutral molecule, bilayer diffusion (PMID 3735402). |
6. Iodide, chloride and H+/K+ transport (#1106)
| Change | Model entities | Evidence |
|---|---|---|
| NIS with 1 Na+ deprecated into the 2 Na+ form; references added to MAR08924 | MAR06989 (iodide + 1 Na+) -> MAR08924 (iodide + 2 Na+, SLC5A5) | NIS is electrogenic with 2 Na+ per I- (PMID 9341168, rat NIS in oocytes; PMID 12488351; UniProt Q92911, RHEA:71207). A 1:1 ratio has not been reported. |
| Pendrin 1 Cl- : 1 I- (before: 1 Cl- : 2 I-) | MAR08671 (SLC26A4) | Electroneutral Cl-/I- exchange: RHEA:72379; PMID 15155570 (pendrin iodide/chloride exchange); PMID 18565999 (mouse Slc26a4, electroneutral). UniProt O43511. |
| H+/K+-ATPase pumps its proton out. Before: ATP[c] + H2O[c] + K+[e] => ADP[c] + Pi[c] + K+[c] + H+[c]; after: ... + H+[e]. Renamed "H+/K+-exchanging ATPase", Rhea master RHEA:22044 | MAR08892 ((ATP4A and ATP4B) or (ATP12A and ATP4B)) | UniProt P20648 (ATP4A) and P54707 (ATP12A): K+(out) + ATP + H2O + H+(in) = K+(in) + ADP + Pi + 2 H+(out). Non-gastric ATP12A activity, PMID 11341842 (insect cells). Human ATP4A gene, PMID 2160952. The old form released the pumped proton into the cytosol. |
| Merged into MAR08892 | MAR01444 "H+/K+ Gastric/Non-Gastric P-ATPase" (2 Na+ out, 1 K+ in, ATP12A and ATP4B) | No H+/K+-ATPase has this stoichiometry. |
Open point: ATP12A also has a Na+/K+ mode (3 Na+ : 2 K+, RHEA:18353), which is not represented after the merge. ATP12A also pairs with ATP1B1/ATP1B3, which the MAR08892 GPR does not include.
7. 2-hydroxyglutarate transport (#1106)
With the split of L- and D-2-hydroxyglutarate, their plasma-membrane transports are now: OAT4 (SLC22A11) exchange with estrone sulfate for (S)-2HG (MAR11733, before a gene-less uniport) and (R)-2HG (MAR20211); OAT1 (SLC22A6) exchange with 2-oxoglutarate for (S)-2HG (MAR11731, existing) and (R)-2HG (MAR20208, new); and NaDC3 (SLC13A3) with 3 Na+ for (S)- and (R)-2HG (MAR20209, MAR20210). Evidence: PMID 18365245 (OAT1 and OAT4 transport glutarate derivatives, including D- and L-2HG) and PMID 21264516. Details and the open point on a mitochondrial 2HG carrier are in #1110.
Implemented in #1100 (merged), #1105 and #1106. Corrections in #1105 to recently curated transports are documented where that curation was made: succinate export (#424), vesicular secretion and BGT1 (#347), Ca2+ channels (#1001) and PXMP2 (#1100).
Activity
edkerk commentedon Oct 5, 2026
Fixed in Human 2.1.0.