Duplicate Bile Acid Synthesis Reactions

[Devlin-Moyer]

Background:

There are currently pathways for converting 25(R)THCA-CoA to cholyl-CoA/cholate in both the [m] and [x] comparments, but these reactions only occur in peroxisomes (sources: (1, 2, 3)), so all of the [m] reactions (and their associated metabolites, since they do not participate in any other reactions) should be removed. Some of these reactions were already mentioned in #634, so I've updated the proposed changes in #634 to skip these reactions, since I have now determined that they do not occur in mitochondria

Diagram with metabolite/reaction IDs:

with metabolite names:

MAR01458 sort of duplicates MAR01642+MAR01646, except there's also conversion of O₂ to water, and I have no idea where that's coming from; also the only gene associated with it, HSD17B4, is only known to be NAD-dependent and not NADP-dependent (Uniprot), so I'm pretty sure it should be removed.

According to the reference associated with MAR01656, "all newly formed primary bile acid-CoA esters within the peroxisomes are instantly conjugated in situ with glycine or taurine before secretion", so MAR01656 should also be removed, because MAR01659 already represents ATP-dependent import of choloyl-CoA into peroxisomes.

While there are enzymes that produce choloyl-CoA in the cytosol, MAR07758, MAR01662, and MAR01666 are also suspicious because:

• The only gene associated with MAR07758, DBI (ENSG00000155368), is not known to be localized to mitochondria, and MAR07758 has no references associated with it
• The references (1, 2, 3, 4) associated with MAR01662 all provide evidence of this reaction occuring in peroxisomes, not mitochondria
• The reference associated with MAR01666 says that "a transport system similar to the carnitine/carnitine palmitoyl transferase machinery used to import fatty acids into the mitochondria may exist to shuttle bile acid intermediates between intracellular compartments; however, no covalently modified derivatives of bile acid intermediates that could represent a transport form have been reported", and neither associated gene (SLC27A2 and SLC27A5) is known to localize to either mitochondrial membrane or transport cholate

Proposed Changes:

• Remove MAR01630, MAR01635, MAR01637, MAR01644, MAR01651, MAR01653, MAR03754, and MAR03760 for being redundant with peroxisomal versions
• Remove MAM00616m, MAM00618m, MAM00749m, MAM00036m, MAM00748m, MAM03367m, MAM01445m, and MAM01514m (MAM03368m was already removed in Fix duplicate metabolites identified by EHMN IDs #649)
• Remove MAR01656 for being redundant with MAR01659
• Remove MAR07758, MAR01662, and MAR01666 for being redundant with peroxisomal versions
• Remove MAR01458 for being redundant with MAR01642+MAR01646

[Devlin-Moyer]

I think it's pretty evident from these diagrams that MAR03764, MAR03756, MAM03368x, and MAM003367x should also be removed for being redundant with MAR01646, MAR01642, MAM00748x, and MAM00036x, respectively, but I didn't include those proposed changes here, since it seems like #616 is going to handle that

[haowang-bioinfo]

nice plot!

it seems that enough evidence has been provided for removing the mitochondria version

Some of these reactions were already mentioned in #634, so I've updated the proposed changes in #634 to skip these reactions, since I have now determined that they do not occur in mitochondria

very good that you noticed overlap issue to this one (please always check overlap/conflict issues before making new ones).

I think it's pretty evident from these diagrams that MAR03764, MAR03756, MAM03368x, and MAM003367x should also be removed for being redundant with MAR01646, MAR01642, MAM00748x, and MAM00036x, respectively, but I didn't include those proposed changes here, since it seems like #616 is going to handle that

A likely sequence of implementation might be: 1) #616; 2) #637; 3) #634. what do you think?

[Devlin-Moyer]

A likely sequence of implementation might be: 1) #616; 2) #637; 3) #634. what do you think?

Sounds reasonable to me

[haowang-bioinfo]

by looking into the pair of MAR01635 versus MAR01632, they both are supported by the annotation of associated gene AMACR (ENSG00000242110) in catalytic activity and subcellular localization. Should MAR01635 really be removed? any more evidence for the deletion?

[Devlin-Moyer]

Yes, because while AMACR is localized to both compartments, the substrate of MAR01635 is not produced in mitochondria, since it is strictly an intermediate in bile acid production, which only occurs in peroxisomes, because only peroxisomal fatty acid oxidation machinery can handle bile acids (sources: 1, 2, 3)

[haowang-bioinfo]

Ok let's move forward according to the provided evidence.

If new evidence supporting existence of mitochondrial bile acid synthesis pathway appeared later, these changes can be easily reverted given the clear history in #637

[haowang-bioinfo]

fixed in #637