#10634 OTR: terms related to fatty acid sythesis

[Jim Hu]

I'm trying to help a CACAO student with annotation from this paper: http://jb.asm.org/content/194/19/5171.full.pdf

This led me to look at terms related to fatty acid synthesis and I am thinking that some rearrangement is needed. The immediate problem is related to:

GO:0033818 ! beta-ketoacyl-acyl-carrier-protein synthase III activity
def: Catalysis of the reaction: acetyl-CoA + malonyl-[acyl-carrier protein] = acetoacyl-[acyl-carrier protein] + CoA + CO2." [EC:2.3.1.180]

EC has 3 kinds of beta-ketoacyl-acyl-carrier-protein synthase activities. The type III activity is involved in priming FA synthesis in bacteria and plants, and catalyzes the reaction used by the protein in the paper. At the level of GO, this is probably the most appropriate term, as it captures the reaction used. The type III activity is associated with a KASIII family of domains. However, the key conclusion of the paper is:
"We herein provide evidence that P. aeruginosa does not utilize a FabH KASIII ortholog as the primary means for FAS initiation like other bacteria studied thus far but instead uses the unannotated open reading frame (ORF) PA5174. …
The identification of the functional role for PA5174, the inaugural member of a highly divergent class of KASI/II domain enzymes, defines the new FabY class of 􏰀-ketoacyl synthase condensation enzymes present in P. aeruginosa."
So, the interesting point is that classification of activities is imperfectly correlated with classification by orthology. I think this term is OK, but it needs a note, which I will add to GONUTS if others agree that it is correct to make that point about activity vs structural classification. The alternative would be a NTR to subclass KASIII activity.

This led me to look at the other KAS terms:
GO:0004315 ! 3-oxoacyl-(acyl-carrier-protein) synthase activity
GO:0033817 ! beta-ketoacyl-acyl-carrier-protein synthase II activity

All three use Malonyl ACP as one substrate. The distinction between GO:0033818 (KASIII) and the others is that KASIII can use free Ac-CoA to prime FA synth, while the others use a second Acyl-ACP. KASIII is only used for priming FA synth, while KASI/II are used for elongation.

Note that the term names for KASII and KASIII include the reaction classification, but the term name for KASI does not, and Kas II and III use beta-ketoacyl while KASI uses 3-oxoacy. These should be consistent, IMO. So, I recommend term name change:

GO:0004315 ! beta-ketoacyl-(acyl-carrier-protein) synthase I activity

I think an intermediate NTR for KAS activity should group the three activities.
GO:xxx ! beta-ketoacyl-(acyl-carrier-protein) synthase activity
def: Catalysis of the reaction: activated C(n) acyl substrate + malonyl-[acyl-carrier protein] = C(n+2) acyl-[acyl-carrier protein] + CoA + CO2." [EC:2.3.1.180]

If I understand it correctly the animal system uses the same chemical activity in a multidomain protein that has KAS domains, ACP domains etc. While the plant and bacterial systems separate these activities into distinct subunits. I would argue that the similarity in the chemistry of using malonyl-ACP vs malonyl-ACP-domain should group these activity terms in GO based on the steps catalyzed rather than whether it's a single chain or multiple chains that perform the activities.

The definitions for KASI and KASII appear to be lifted from EC, but the catalysis of reaction definitions don't capture the relevant distinctions (I think).

Finally I would argue that GO:0004312 ! fatty acid synthase activity should be obsoleted, as this is a group of enzymatic activities and is thus a process rather than a molecular function.

[Paola Roncaglia]

In case it's any help, we also have an open ticket about fatty acid biosynthesis and elongation:

https://sourceforge.net/p/geneontology/ontology-requests/10586/

[Jim Hu]

Thanks… not that the other ticked is about the process terms, while here we are looking at activities. But I may weigh on that one too.

[Peter D\'Eustachio]

Don't GO molecular_function terms strictly follow function, so when that diverges from structure / orthology, the annotation should be to the function? Indeed, structure and function do diverge sometimes and if GO annotations can be kept strictly functional while InterPro domains stay strictly structural, a phylogeny of these divergences pops out - possibly interesting.

[David Hill]

The function terms should only reflect a function independent of how that function arose. That said, I think we should have a look at the consistency issues brought up in this item and try to make some improvements on this in both function and process. Perhaps we can take a bit of time to look at this together at the upcoming meeting. It might be a candidate for a new biochemical pathway/metabolism project.

[Paola Roncaglia]

Hi, in case it's helpful, there's an old SF ticket about
GO:0004321 fatty-acyl-CoA synthase activity
and
GO:0004312 fatty acid synthase activity:
https://sourceforge.net/p/geneontology/ontology-requests/7887/
It was assigned to Amelia. David, perhaps you could take it to go with this one, or close the old one if there's any duplication?

Many thanks,
Paola

[Paola Roncaglia]

Also see https://sourceforge.net/p/geneontology/ontology-requests/7887/