ubiquinone biosynthesis

[ValWood]

• Gene List

Systematic ID	Gene name	Product description	Budding yeast orthologs	Human orthologs
SPBC146.12	coq6	2-octaprenyl-6-methoxyphenol hydroxylase Coq6	COQ6	COQ6
SPBC337.15c	coq7	3-demethoxyubiquinol 3-hydroxylase 3-demethoxyubiquinone 3-hydroxylase (NADH) activity Coq7	CAT5	COQ7
SPBC2D10.18	coq8	ABC1 kinase family ubiquinone biosynthesis ATPase Coq8	COQ8	COQ8A,COQ8B
SPCC4G3.04c	coq5	C-methyltransferase	COQ5	COQ5
SPAC19G12.12	dlp1	decaprenyl diphosphate synthase subunit 2 Dlp1	COQ1	PDSS2
SPBPJ4664.01	dps1	decaprenyl diphosphate synthase subunit Dps1/Coq1	COQ1	PDSS1
SPCC162.05	coq3	decaprenyldihydroxybenzoate methyltransferase Coq3	COQ3	COQ3
SPAC1556.03	azr1	mitochondrial protein phosphatase Azr1	PTC7	PPTC7
SPAC1071.11	coq12	NAD reductase Coq12
SPAC56F8.04c	ppt1	para-hydroxybenzoate--polyprenyltransferase Ppt1	COQ2	COQ2
SPCC1840.09	coq11	ubiquinone biosynthesis protein Coq11	COQ11	NDUFA9
SPAC1687.12c	coq4	ubiquinone biosynthesis protein Coq4	COQ4	COQ4
SPAC19G12.11	coq9	ubiquinone biosynthesis protein Coq9	COQ9	COQ9

• Noctua Model

http://noctua.geneontology.org/workbench/noctua-visual-pathway-editor/?model_id=gomodel%3A662af8fa00000408
http://noctua.geneontology.org/workbench/noctua-alliance-pathway-preview/?model_id=gomodel%3A662af8fa00000408

• Tickets
  3-demethoxyubiquinone 3-hydroxylase (NADH) activity geneontology/go-ontology#29305
  ubiquinone biosynthesis process clean up geneontology/go-ontology#29313

• Tasks Pombase

• Detect orthologs for Coq12?
• Find out how coq11 & coq12 fit into the pathway (I don't think this is known)
• Update PomBase products lines for Coq4 and Coq9 and Coq8
• coq4 needs correct activity 3-demethoxyubiquinone 3-hydroxylase (NADH) activity geneontology/go-ontology#29305 (not sure what it is precisely)
• represent dpl1,2 as complex to align with FB

[hattrill]

• @ValWood think human PDSS2 is a better ortholog for dlp1
  fixed ortholog

[hattrill]

Orthologs for sp genes - all have GO ubiquinone biosynthesis annotations:

Systematic ID	Gene name	Product description	Budding yeast orthologs	Human orthologs	D.mel FBgn	D.mel symbol
SPBC146.12	coq6	2-octaprenyl-6-methoxyphenol hydroxylase Coq6	COQ6	COQ6	FBgn0031713	Coq6
SPBC337.15c	coq7	3-demethoxyubiquinol 3-hydroxylase Coq7	CAT5	COQ7	FBgn0029502	Coq7
SPBC2D10.18	coq8	ABC1 kinase family ubiquinone biosynthesis ATPase Coq8	COQ8	COQ8A,COQ8B	FBgn0052649	Coq8
SPCC4G3.04c	coq5	C-methyltransferase	COQ5	COQ5	FBgn0030460	Coq5
SPAC19G12.12	dlp1	decaprenyl diphosphate synthase subunit 2 Dlp1	COQ1	PDSS2	FBgn0037044	Pdss2
SPBPJ4664.01	dps1	decaprenyl diphosphate synthase subunit Dps1/Coq1	COQ1	PDSS1	FBgn0051005	qless
SPCC162.05	coq3	decaprenyldihydroxybenzoate methyltransferase Coq3	COQ3	COQ3	FBgn0032922	Coq3
SPAC1556.03	azr1	mitochondrial protein phosphatase Azr1	PTC7	PPTC7	FBgn0035228/FBgn0039694/FBgn0029949	CG12091/fig/CG15035
SPAC1071.11	coq12	NAD reductase Coq12
SPAC56F8.04c	ppt1	para-hydroxybenzoate--polyprenyltransferase Ppt1	COQ2	COQ2	FBgn0037574	Coq2
SPCC1840.09	coq11	ubiquinone biosynthesis protein Coq11	COQ11	NDUFA9	FBgn0037001	ND-39
SPAC1687.12c	coq4	ubiquinone biosynthesis protein Coq4	COQ4	COQ4	FBgn0052174	Coq4
SPAC19G12.11	coq9	ubiquinone biosynthesis protein Coq9	COQ9	COQ9	FBgn0050493	Coq9

[hattrill]

From Invertebrate models for coenzyme q10 deficiency

Check:
-Where does COQ10A/COQ10B fit in? For COQ10A, UniProt says 'Required for the function of coenzyme Q in the respiratory chain. May serve as a chaperone or may be involved in the transport of Q6 from its site of synthesis to the catalytic sites of the respiratory complexes', so does not belong to 'ubiquinone biosynthesis'. In D.mel, the ortholog is CG9410.
-heix (human ortholog UBIAD1) 'ubiquinone biosynthetic process via 3,4-dihydroxy-5-polyprenylbenzoate' (IBA)
-Rtn4ip1 'ubiquinone biosynthetic process' (IMP) ; 'NADPH dehydrogenase (quinone) activity' (ISS) Mitochondrial matrix RTN4IP1/OPA10 is an oxidoreductase for coenzyme Q synthesis
-CG15035/CG12091 regulation of ubiquinone biosynthetic process (IBA) (human ortholog PPTC7) - paralogs for fig which also has same IBA - so should be included on table

[ValWood]

Agreed about COQ10, it is the carrier which delivers to the etc,
https://www.pombase.org/gene/SPCC16A11.07
I ISSd from https://www.yeastgenome.org/locus/S000005368#go
but it shouldn't be metabolism.
What process would we annotate too? In some ways its a problem that we can't describe a pathway in a way which includes metabolism and transport....

[ValWood]

Rtn4ip1 , interesting, I have this annotated as an unknown with https://www.pombase.org/gene/SPBC16A3.02c
mitochondrial CH-OH group oxidoreductase, human RTN4IP1 ortholog, implicated in isoprenoid or sterol metabolism
In yeast it also seems to be detected in the ER as well as the mitochondrion (and reticulon the interacting partner is involved in ER membrane bending)

[hattrill]

heix/UBIAD1 is involved in non-mitochondrial Q10 synthesis: "Ubiad1 is an antioxidant enzyme that regulates eNOS activity by CoQ10 synthesis The rate-limiting enzyme for the biosynthesis of CoQ10 is the enzyme that catalyzes the condensation of the polyisoprenoid chain with the benzoquinone ring. So far, the mitochondrial COQ2 enzyme has been considered the only prenyltransferase able to catalyze this reaction (Trevisson et al., 2011). Here, we identified UBIAD1 as a vertebrate CoQ10 prenyltransferase. UBIAD1 contains an UbiA prenyltransferase domain also present in vertebrate COQ2. Although COQ2 encodes a mitochondrial prenyltransferase, we found that UBIAD1 resides in the Golgi compartment where it produces CoQ10. While the presence of CoQ10 in nonmitochondrial membranes was previously explained by the existence of specific mechanisms for its redistribution within the cell (Crane and Morre, 1977; Jonassen and Clarke, 2000), our data now formally demonstrate that CoQ10 are synthetized in the Golgi compartment.
...In favor of our hypothesis of a Golgi-synthetized CoQ10, it has been recently reported that COQ6, COQ7, and COQ9, which are critical enzymes for CoQ10 maturation, are also localized in the Golgi compartment (Heeringa et al., 2011)."

[hattrill]

Agreed about COQ10, it is the carrier which delivers to the etc, https://www.pombase.org/gene/SPCC16A11.07 I ISSd from https://www.yeastgenome.org/locus/S000005368#go but it shouldn't be metabolism. What process would we annotate too? In some ways its a problem that we can't describe a pathway in a way which includes metabolism and transport....

I am not sure whether this would or could go under 'respiratory electron transport chain', I see that the interpro entry links to 'cellular respiration'....Reactome models it like this:

[ValWood]

Chat GPT didn't seem convinced, it would be nice to see some follow up:

You're correct that some studies suggest enzymes involved in the biosynthesis of Coenzyme Q10 (ubiquinone) are found in the Golgi apparatus, and this adds nuance to the classic understanding of its synthesis. Here's a more detailed perspective:

Classical View:

The majority of CoQ10 synthesis occurs in the mitochondria, where enzymes like COQ6, COQ7, and COQ9 are localized. These enzymes are involved in modifying the benzoquinone ring and attaching the isoprenoid tail.
Golgi Localization:

Research has identified certain CoQ-related enzymes, such as COQ8B (ADCK4), that may localize partially to the Golgi apparatus in some cell types.
The Golgi could provide a platform for certain modification or assembly steps, or it might participate in trafficking components like the prenyl chain intermediates.
Interplay Between Organelles:

CoQ10 synthesis likely involves inter-organelle cooperation:
The mitochondria handle core enzymatic reactions.
The endoplasmic reticulum (ER) and Golgi may contribute by modifying lipid precursors or processing intermediates.
Context of Findings:

Experimental observations about Golgi-localized CoQ enzymes could vary depending on the cell type or experimental method, including overexpression studies or fluorescent tagging that might alter protein localization.
Thus, while mitochondria remain the central hub for ubiquinone synthesis, emerging evidence points to possible roles for the Golgi and other organelles in the process, especially in specific cellular contexts or for enzyme localization.

[hattrill]

For UBIAD1, seems to be involved in vitamin K2 subtype menaquinonen-4 synthesis and there are a handful of papers that suggest a role in non-mitochondrial Q10 synthesis. "UBIAD1 catalyzes transfer of the 20-carbon geranylgeranyl group from geranylgeranyl pyrophosphate to menadione (vitamin K3 ) derived from plant-derived phylloquinone (vitamin K1 ), generating MK-4(menaquinone-4, vitamin K 2) (Figure 1) (Nakagawa et al., 2010; Hirota et al., 2013). It has also been
proposed that UBIAD1 mediates polyprenylation of 4-hydroxybenzoate to produce 3-polyprenyl-4-
hydroxybenzoate (PPHB), an intermediate in the synthesis of CoQ10 (coenzyme Q10 or ubiquinone-10)"

The prenyltransferase UBIAD1 is the target of geranylgeraniol in degradation of HMG CoA reductase
Synthesis, function, and regulation of sterol and nonsterol isoprenoids
Cancer cell stiffening via CoQ10 and UBIAD1 regulates ECM signaling and ferroptosis in breast cancer

So minded to leave the IBAs as they are, just not add heix/UBIAD1 to the 'cononical' mitochodrial ubiquinone biosynthesis set here.

[ValWood]

so according to Rhea https://www.uniprot.org/uniprotkb/?query=rhea:44564
it's the same reaction as coq2
Interesting if this is true but this is an early pathway step so it would either need to tranlocate to the mitochondria, or the full pathway would need to be in the ER ( 4-hydroxybenzoate is a very early intermediate). I added it to my model as it keeps cropping up.

[hattrill]

For flies, heix/UBIAD1 is localized mainly to the ER but a small proportion is mitochondrial in some tissues. There is also a suggestion that its role in the mitochondrion might be vitamin K2. I'll have to have a good think about where and what to do with it!

[hattrill]

Disputed PMID:30267671 fig - regulation of ubiquinone biosynthetic process - should be for the paralog CG12091 and asked to be changed to positive regulation like s.p. azr1

fig and CG15035 are testis-specific
CG12091 - is ubiquitously expressed so use in model

[hattrill]

Just noting where I've got to with annotating and modeling this.

Noctua Model
Sheet for tracking work

Main points:

1. For D.mel, there are 3 species of coenzyme Q: Q8-Q10, the major species is Q9. These chain lengths have been detected as intermediates in the pathway and so I conclude that they are not just dietary source.
2. The variance in chain length probably results from the polyprenyl diphosphate synthase complex having a wider 'window' for releasing the chain - but I am just guessing. Certainly, the species-specific variation of coenzyme Q is attributed to this enzyme. The starting point seems to be farnesyl diphosphate from Fpps (I have not found any evidence that the acceptor is different, but I will note that I cannot say that this is absolutely true). Also, add a note to say that Fpps has been shown to also localise to the mitochondrion PMID:17198737
   

Suggested biosynthesis route for nonadiphosphate
Step-by-Step Chain Elongation, starting with farnesyl diphosphate (C15):
Initial substrate with 3 isoprene units (C15).
Add IPP units sequentially:
First addition: C15 + C5 → C20 (Geranylgeranyl Diphosphate, GGPP).
Second addition: C20 + C5 → C25 (Geranylfarnesyl Diphosphate).
Third addition: C25 + C5 → C30.
Fourth addition: C30 + C5 → C35.
Fifth addition: C35 + C5 → C40.
Sixth addition: C40 + C5 → C45 (Nonaprenyl Diphosphate).
Final product:
Nonaprenyl Diphosphate (C45) has 9 isoprene units.
3. As the ubiquinone biosynthesis pathway can tolerate different chain lengths and there is really no discernable difference for the other pathway enzymes, need chain length agnostic MF terms to model D.mel pathway. Steven has made ticket to fix BP: geneontology/go-ontology#29419
4. As other groups have been working on this pathway, there are improvements that will trickle down soon, so just have to wait a bit. e.g. need a chain agnostic term for coq6 step see geneontology/go-ontology#29313
5. Steven has reviewed the MFs and fixed various aspects here geneontology/go-ontology#29420
6. Reactome model is recent and is well-reviewed, used as guide. Similar to Metacyc (human). There is a decarboxylase, and a hydrolase step that are still not properly attributed to enzymes - Coq4 seems to be involved but the exact input and outputs are debated. I have modeled this very loosely. The top flow of the diagram below is the one consistent with the models in other resources:

7. Although ubiquinone (the oxidized form) is the one that supplies the respiratory chain, ubiquinone (Q) and ubiquinol (QH2) probably exist in equilibrium.

Models tend to show the reduced form, and so to be consistent, each step in the pathway should be the "-ol" version rather than the "-one" version. There is an inconsistency on the Coq5 step and Steven is following this up.

[ValWood]

I have also loosely modelled coq4. There is a paper referred to in
geneontology/go-ontology#29305 should help us here.
which seems useful for this step PMID:38425362 but I still couldn't figure out the precise activity

@Antonialock was going to do ubiquinone pathway but mentioned @sylvainpoux is doing it : Sylvain this might be useful for you.

[hattrill]

The reactome pathway is really recently reviewed. For this step they note:
COQ4 is a zinc-dependent enzyme that catalyzes the decarboxylation step in CoQ biosynthesis (Pelosi et al., 2024 and Nicoll et al., 2024). However, the substrate and the product of the decarboxylation step remain ambiguous since Nicoll et al. reported a decarboxylation of 3-methoxy-4-hydroxy-5-decaprenylbenzoate (MHDB) into 2-methoxy-6-decaprenylphenol (DMPhOH), whereas Pelosi et al. proposed that COQ4 catalyzes an oxidative decarboxylation which yields a hydroxylated product. In this case, DHB rather than MHDB was proposed a substrate for COQ4, generating 3-polyprenyl-4-hydroxyphenol as a product. We are provisionally showing the first reaction until more data is available.

[sylvainpoux]

Hi,
we plan to make a new GO-CAM pathway for the ubiquinone biosynthesis pathway in human based on the work of Nicoll et el. 2024. Anne and Kristian from Rhea do not believe that the dual activity of COQ4 described by Pelosi et al. is convincing.

Thanks

Sylvain

[hattrill]

That's great @sylvainpoux would be good to align our model with yours, especially as Anne and Kristain have been working on the reactions.

[ValWood]

Well I'm pleased to report that the pombase and flybase models are identical (topologically)

Flybase

PomBase

[sylvainpoux]

Hi @ValWood and @hattrill,

I'm starting to work on the GO-CAM ubiquinone pathway and would have a bunch of comments/questions:

1. According to Nicoll et al. (PMID:38425362), COQ6 mediates 2 non-consecutive steps in the ubiquinone pathway (see the preliminary COQ6 human entry below).

I noticed that you only curated the first step (C5-ring hydroxylation) and not the other one (C1-hydroxylation downstream of COQ4). Is there a reason for this?

2. I noticed that you added COQ8 as a protein kinase for COQ3, I however think that it is still speculative. Moreover, we don't know whether it would affect the first or second step mediated by COQ3. I planned to not add COQ8 (COQ8A/COQ8B) in my GO-CAM model.

3. A question for you Helen, do you know the length of ubiquinone in Drosophila?

4. Last but not least, in human and yeast, we know the ferredoxin/ferredoxin reductase (FDX2/FDXR and YAH1/ARH1, respectively) that provides electrons to COQ6. I will add FDX2 and FDXR in the human GO-CAM model.

Thanks

Sylvain

COQ6 human Q9Y2Z9
CC FUNCTION: FAD-dependent monooxygenase required for two non-consecutive steps during ubiquinone biosynthesis. {EXP|Ref.10, EXP|Ref.13}· Required for the C5-ring hydroxylation during ubiquinone biosynthesis by catalyzing the hydroxylation of 4-hydroxy-3-(all-trans-decaprenyl)benzoic acid to 3,4-dihydroxy-5-(all-trans-decaprenyl)benzoic acid. {EXP|Ref.10, EXP|Ref.13}· Also acts downstream of COQ4, for the C1-hydroxylation during ubiquinone biosynthesis by catalyzing the hydroxylation of 2-methoxy-6-(all-trans-decaprenyl)phenol to 2-methoxy-6-(all-trans-decaprenyl)benzene-1,4-diol. {EXP|Ref.13}· The electrons required for the hydroxylation reaction are funneled indirectly to COQ6 from NADPH via a ferredoxin/ferredoxin reductase system composed of FDX2 and FDXR. {EXP|Ref.10, EXP|Ref.13}
CC CATALYTIC ACTIVITY
Reaction=4-hydroxy-3-(all-trans-decaprenyl)benzoate + 2 reduced [2Fe-2S]-[ferredoxin] + O2 + 2 H(+) = 3,4-dihydroxy-5-(all-trans-decaprenyl)benzoate + 2 oxidized [2Fe-2S]-[ferredoxin] + H2O [Rhea:81259]
Evidence={MODM|HAMAP-Rule:MF_03193, EXP|Ref.10, CUR|Ref.13}
CC CATALYTIC ACTIVITY
Reaction=2-methoxy-6-(all-trans-decaprenyl)phenol + 2 reduced [2Fe-2S]-[ferredoxin] + O2 + 2 H(+) = 2-methoxy-6-(all-trans-decaprenyl)benzene-1,4-diol + 2 oxidized [2Fe-2S]-[ferredoxin] + H2O [Rhea:81295]
Evidence={MODM|HAMAP-Rule:MF_03193, CUR|Ref.13}

[hattrill]

Hi @sylvainpoux - the major species of ubiquinone in D.mel is Q9, but Q8 (∼5 %) and CoQ10 (∼13 %) also exist Endogenous coenzyme Q content and exogenous bioavailability in D. melanogaster.

Although I can point to no direct assay, it seems likely that this variation occurs at the polyprenyl diphosphate synthase step, as others have reported seeing Q8-10 length intermediates in D.mel pathway mutants and polyprenyl diphosphate synthase is pointed to as the reason for chain length variation between species.

This is why I want to model using chain-length agnostic terms.

[hattrill]

I cross checked my model with the reactome model . For the second COQ6 step this is the note they added:
"The enzymatic conversion of 2-methoxy-6-decaprenylphenol (DMPhOH) to 2-methoxy-6-decaprenyl-1,4-benzoquinol (MDMQ10H2) is inferred from the equivalent reaction in yeast (Gin et al. 2003, Ozeir et al. 2011). . It was thought at one time that the flavin dependent monooxygenase, COQ6, was the enzyme that catalysed this reaction, however, it has been subsequently shown that COQ6 is not essential for this reaction (Ozeir et al. 2011). Yet, Nicoll et al. reported low levels of in vitro activity for a resurrected ancestral COQ6, opening the possibility that COQ6 might be a bifunctional hydroxylase operating on carbon C5 and C1 of CoQ precursors (Nicoll et al., 2024). Note that the proposal by Pelosi et al. that COQ4 catalyzes an oxidative decarboxylation eliminates reaction from the CoQ biosynthesis pathway (Pelosi et al., 2024) is an alternative hypothesis."

So I have modelled a gap for this and added a very non-commital placeholder for COQ4 before this as reactome indicated two possible rxns for this.

[sylvainpoux]

Hi @ValWood and @hattrill

While reviewing the ubiquinone pathway in human, I requested a bunch of new terms and an update of one existing term
You might be interested to have a look to the ticket
geneontology/go-ontology#29459
Basically, some reactions were missing. Please note that these reactions have been accepted by the enzyme commission and are associated with EC numbers and Rhea reactions

Thanks

Sylvain

[hattrill]

Thanks @sylvainpoux! That's great.

[sylvainpoux]

Hi @ValWood and @hattrill

I just finished the ubiquinone pathway in human.

Thanks

Sylvain

[ValWood]

Hi Sylvain,

Thanks for this. I revised my model to replace the coq4 placeholder with coq4/coq6 and the correct/new activities that you curated from the new paper PMID:38425362

I added the unknown mono-oxygenase, but then I was confused because the inputs and outputs already matched up in my model.

So I
PomBase Model, I had

coq5
2-methoxy-6-polyprenyl-1,4-benzoquinol methyltransferase activity (GO:0008425)
Definition
Catalysis of the reaction: a 2-methoxy-6-all-trans-polyprenyl-1,4-benzoquinol + S-adenosyl-L-methionine = a 6-methoxy-3-methyl-2-all-trans-polyprenyl-1,4-benzoquinol + S-adenosyl-L-homocysteine + H+.

provides input for

coq7
-demethoxyubiquinol 3-hydroxylase Coq7 (not that this is a monooxygenase)
Catalysis of the reaction: a 6-methoxy-3-methyl-2-all-trans-polyprenyl-1,4-benzoquinol + donor-H2 + O2 = acceptor + a 3-demethylubiquinol + H2O.

You have

UniProt HUMAN model
coq5
2-methoxy-6-polyprenyl-1,4-benzoquinol methyltransferase activity (GO:0008425)
Definition
Catalysis of the reaction: a 2-methoxy-6-all-trans-polyprenyl-1,4-benzoquinol + S-adenosyl-L-methionine = a 6-methoxy-3-methyl-2-all-trans-polyprenyl-1,4-benzoquinol + S-adenosyl-L-homocysteine + H+.

input. 2-decaprenyl-6-methoxy-3-methylhydroquinone
unknown monooxygenase
output (6-methoxy-3-methyl-2-all-trans-polyprenyl-1,4-benzoquinol) this is Ubiquinol

coq7
demethoxyubiquinol 3-hydroxylase Coq7
Catalysis of the reaction: a 6-methoxy-3-methyl-2-all-trans-polyprenyl-1,4-benzoquinol + donor-H2 + O2 = acceptor + a 3-demethylubiquinol + H2O.
according to ChatGPT
2-methoxy-6-all-trans-polyprenyl-1,4-benzoquinol == 2-decaprenyl-6-methoxy-3-methylhydroquinone
So I don't think the additional unknown monooxygenase is required? Note that coq7 IS a mono-oxygenase ignore

What do you think?

val

[ValWood]

Note to self, I also need to add the ferrioxin to the PomBase model when the annotations come through

[sylvainpoux]

Hi @ValWood,

My model is based on PMID:38425362 and if you look to the model you have:

1. COQ5 has output 2-decaprenyl-6-methoxy-3-methylhydroquinone (CHEBI:64181)

2. A protein: substrate: 2-decaprenyl-6-methoxy-3-methylhydroquinone (CHEBI:64181) and product: 2-methoxy-5-methyl-6-all-trans-polyprenylbenzoquinone (CHEBI:231829)

3. COQ7 substrate: 2-methoxy-5-methyl-6-all-trans-polyprenylbenzoquinone (CHEBI:231829)and product: 3-demethylubiquinone(1-) [CHEBI:231825)

So I guess we need an additional protein (which could be RTN4IP1, but we need additional evidence)

There is an error in your reaction for COQ7, the correct reaction/GO is: Catalysis of the reaction: a 5-methoxy-2-methyl-3-(all-trans-polyprenyl)benzoquinone + NADH + O2 = a 3-demethylubiquinone + NAD+ + H2O. vw FIXED

All structures can be found in PMID:38425362. Could you check with a chemist if ChatGPT is correct? To my knowledge an ubiquinol is not the equivalent of an ubiquinone. ignore

Thanks

Sylvain

[ValWood]

Re>
There is an error in your reaction for COQ7, the correct reaction/GO is: Catalysis of the reaction: a 5-methoxy-2-methyl-3-(all-trans-polyprenyl)benzoquinone + NADH + O2 = a 3-demethylubiquinone + NAD+ + H2O.

I fixed this one. @hattrill this will apply to FB model too. I also query the other term we initially used here:
geneontology/go-ontology#29563
(maybe it is only for prokaryotes?)

[sylvainpoux]

@ValWood

Could you precise who made an error in the reaction fro COQ7?

Sylvain

[ValWood]

Could you precise who made an error in the reaction fro COQ7?

PomBase. PomBase seems to be the experimental source. This is propagated via PAINT so it should fix automatically once geneontology/go-ontology#29563 is addressed

[hattrill]

Hi @ValWood @sylvainpoux

Have updated the fly model based on the points above.

I am having an issue with the FDX/FDXR input (perhaps I am misunderstanding this, so forgive the tangent if so):
In the human model, FDX2 -> FDXR -> COQ6.

But PMID:38425362 suggests that FDXR transfers e- from NADPH to reduce FDX2. FDX2 then transfers e- then reduces COQ6 = FDXR (+2NADPH) -> FDX2 (+2NADP+) -> COQ6

I am also a bit confused about the MF for FDXR -

ferredoxin-NADP+ reductase activity
DEF: Catalysis of the reaction: reduced ferredoxin + NADP+ = oxidized ferredoxin + NADPH + H+.
Comments
Note that this term specifically refers to the reaction proceeding in the direction shown; it should therefore be used to annotate gene products that catalyze the oxidation of reduced ferredoxin or adrenodoxin

i.e. reducing NADP+ and oxidizing ferredoxin

There is a term,
'NADPH-adrenodoxin reductase activity' GO:0015039 which goes in this direction oxidized adrenodoxin + NADPH + H+ => reduced adrenodoxin + NADP+.

adrenodoxin = FDX1 (ferredoxin 1)

but no 'generic' NADPH-ferredoxin reductase activity which we could use for the reduction of FDX2 by FDXR.

Seems a bit strange to distinguish between FDX2 and FDX1 as makes the term based on one gp target. Plus adrenodoxin seems a bit of an archaic term based on process/expression rather than MF.

The GO structure we have at the moment is:

Think that renaming and redefining NADPH-adrenodoxin reductase activity GO:0015039 to NADPH-ferredoxin reductase activity GO:0015039 might solve this issue. (would have to fix xrefs as well)

[ValWood]

I agree about adrenodoxin. I tried to address that it seems to represent a gene product here but abandoned geneontology/go-ontology#27715
Happy to try to resolve again but note that currently, NADPH-adrenodoxin has a taxon restriction that would need to be removed.

I agree that the directionality should be instantiated in the reaction, not in a comment, but your solution of repurposing GO:0015039 seems to work

[hattrill]

I agree about adrenodoxin. I tried to address that it seems to represent a gene product here but abandoned geneontology/go-ontology#27715 Happy to try to resolve again but note that currently, NADPH-adrenodoxin has a taxon restriction that would need to be removed.

I agree that the directionality should be instantiated in the reaction, not in a comment, but your solution of repurposing GO:0015039 seems to work

That sounds like a plan - for GO:0015039 might have to get rhea to look at this, but could use think xref to RHEA:20125as it can go right to left. As this is also an xref for ferredoxin-NADP+ reductase activity GO:0004324, would that make it a 'broad synonym' as an exact one can't be shared by >1 term ???
-Remove taxon constraint
-Re-define as "Catalysis of the reaction: oxidized ferredoxin + NADPH + H+ -> reduced ferredoxin + NADP+.
-Remove note.

And for GO:0004324 ferredoxin-NADP+ reductase activity
Change def to "Catalysis of the reaction: reduced ferredoxin + NADP+ -> oxidized ferredoxin + NADPH + H+."

However,
GO:0008937 ferredoxin-NAD(P) reductase activity
GO:0008860 ferredoxin-NAD+ reductase activity
don't seem to restrict directionality, so perhaps they can stay as they are and even move the renamed/defined GO:0015039 under
GO:0008937 ferredoxin-NAD(P) reductase activity ?

Shall I make a separate ticket for this or re-open geneontology/go-ontology#27715?

[ValWood]

I would make a separate ticket -top of the pile ;), but link to the old one to show that issues have been raised before.