Electron Transport Flavoprotein

[Devlin-Moyer]

Note: significantly edited after discovering that the Metabolic Atlas website showed an older version of MAR06911 with a different GPR

Background

MAR000562 and MAR00563 are redox reactions between Electron Transport Flavoprotein and FAD. The GPR of MAR00562 is ETFA or ETFB, and the GPR of MAR00563 is ETFDH.

Before #489, the GPR of MAR06911 (redox between FAD and ubiquinone) was ETFA or ETFB or ETFDH or SDHC. That PR changed the GPR of MAR06911 to SDHA and SDHB and SDHC and SDHD, with the justification (given in #468) that MAR06911 represented the reaction catalyzed by Complex II/Succinate Dehydrogenase, and ETFA, ETFB, and ETFDH are not subunits of Complex II/SDH.

ETFA and ETFB encode the two subunits of the Electron Transport Flavoprotein heterodimer, which binds to a number of other mitochondrial FAD-dependent enzymes, transfers electrons from their FAD cofactors to its FAD cofactor, then binds to ETFDH in the mitochondrial membrane and passes those electrons onto ETFDH's FAD cofactor, which ETFDH uses to reduce ubiquinone to ubiquinol. (source, source). Here's a diagram (from this paper) that illustrates the situation (specifically the bottom center; Q/QH2 is ubiquinone/ubiquinol):

Problems

The Electron Transport Flavoprotein "metabolites", MAM03573m and MAM03574m, only appear in MAR00562 and MAR00563, so they don't really represent the role of ETF in binding to other FAD-dependent enzymes and transfering electrons between their FAD cofactors. The enzymes that ETF interacts with (shown in the above diagram) collectively catalyze >100 reactions in Human-GEM, most of which involve FAD (i.e. MAM01802m and MAM01803m) and not MAM03573m and MAM03574m. I think changing all of those reactions to involve MAM03573m and MAM03574m instead of FAD would be worse than just returning ETFA, ETFB and ETFDH to the GPR of MAR06911 and removing MAR00562, MAR00563, MAM03573m and MAM03574m.

I also think the Complex II/SDH subunits should be removed from the GPR of MAR06911, since, as discussed in #607, MAR04652 and MAR08743 also represent the reaction catalyzed by Complex II/SDH. The proposed changes in #607 would leave MAR04652 as a single one-step reaction that represents the Complex II/SDH reaction, so leaving the Complex II/SDH subunits in MAR06911 would make it a partial duplicate of MAR04652.

Also, as the above diagram shows, several other FAD-dependent enzymes (namely PRODH, PRODH2, GPD2, DHODH and CHDH) are capable of directly transfering electrons from their FAD/FMN cofactors to ubiquinone without the aid of ETFA/B and ETFDH, so I think they should be added to the GPR of MAR06911.

As a side note, the only enzyme that appears on that diagram that isn't already in Human-GEM is SQOR, which, according to Uniprot, catalyzes reactions that are also not currently in Human-GEM (as far as I can tell), but I figure that discussing whether or not SQOR and any of its associated reactions should be added to Human-GEM can be a separate issue.

Proposed Changes:

• Remove MAR00562 and MAR00563 for being redundant with MAR06911
• Remove MAM03573m and MAM03574m
• Change the GPR of MAR06911 to ETFA or ETFB or ETFDH or PRODH or PRODH2 or GPD2 or DHODH or CHDH, i.e. ENSG00000140374 or ENSG00000105379 or ENSG00000171503 or ENSG00000100033 or ENSG00000250799 or ENSG00000115159 or ENSG00000102967 or ENSG00000016391

[JonathanRob]

I agree completely - including a separate reaction for ETF redox is overkill, and can instead be captured by adding it to the GPR of the relevant reactions.

And I suppose it does make sense to add the relevant FADH-ubiquinone redox enzymes to the GPR of MAR06911.

[haowang-bioinfo]

fixed in #612