glycerol-1-phosphatase activity / glycerol-3-phosphatase activity

[sjm41]

GO currently has these two terms (with the same MetaCyc xref):

id: GO:0000121 name: glycerol-1-phosphatase activity def: "Catalysis of the reaction: glycerol-1-phosphate + H2O = glycerol + phosphate." [EC:3.1.3.21] synonym: "alpha-glycerol phosphatase activity" RELATED [EC:3.1.3.21] synonym: "alpha-glycerophosphatase activity" RELATED [EC:3.1.3.21] synonym: "glycerol 3-phosphatase activity" RELATED [EC:3.1.3.21] synonym: "glycerol 3-phosphate phosphohydrolase activity" RELATED [EC:3.1.3.21] synonym: "glycerol-1-phosphate phosphohydrolase activity" RELATED [EC:3.1.3.21] synonym: "glycerol-3-phosphate phosphatase activity" RELATED [EC:3.1.3.21] xref: EC:3.1.3.21 xref: MetaCyc:GLYCEROL-1-PHOSPHATASE-RXN xref: RHEA:11476 is_a: GO:0016791 ! phosphatase activity

id: GO:0043136 name: glycerol-3-phosphatase activity def: "Catalysis of the reaction: glycerol 3-phosphate + H2O = glycerol + phosphate." [GOC:jl] xref: MetaCyc:GLYCEROL-1-PHOSPHATASE-RXN is_a: GO:0016791 ! phosphatase activity

Note that EC:3.1.3.21 is associated with two reactions: H2O + sn-glycerol 1-phosphate <=> glycerol + phosphate (= RHEA:46084/MetaCyc:RXN-14964) H2O + sn-glycerol 3-phosphate <=> glycerol + phosphate (= RHEA:66372/MetaCyc:RXN-14965)

Then RHEA:11476 (which is the xref on GO:0000121) is the 'general form' of those two specific reactions. It has the reaction: glycerol 1-phosphate + H2O = glycerol + phosphate Not sure why, but RHEA only links EC:3.1.3.21 to the two specific child reactions, and not to RHEA:11476.

So...what to do here in the GO?

Simplest (but maybe naive?) option would be to merge GO:0043136 into GO:0000121 and keep RHEA:11476 and EC:3.1.3.21 as xrefs. (Though the def xref should be changed to RHEA:11476). But I'd change the name from 'glycerol-1-phosphatase activity' to 'glycerol phosphate phosphatase activity', and maybe 'glycerol-1-phosphate' should also be changed to just 'glycerol phosphate' in the def?

The alternative option might be to have a little hierarchy like this:

glycerol phosphate phosphatase activity (GO:0000121) = EC:3.1.3.21/RHEA:11476/MetaCyc:GLYCEROL-1-PHOSPHATASE-RXN
    |_sn-glycerol 1-phosphate phosphatase activity (NTR) = RHEA:46084/MetaCyc:RXN-14964
    |_sn-glycerol 3-phosphate phosphatase activity (GO:0043136) = RHEA:66372/MetaCyc:RXN-14965

(Note RHEA still refers to 'glycerol 1-phosphate' in their generic/parental reaction, so I'm not really sure if it's correct to drop the '1' from the parental GO term name or not.....)

I'm coming at all this after encountering this activity in PMID:26755581 (human) and PMID: 35017476 (worm).

Advice from @kaxelsen would be appreciated!

[kaxelsen]

In Rhea we prefer to link the unambiguous reactions (especially when there only are 2 alternatives). glycerol 1-phosphate is ambiguous as it includes sn-glycerol 1-phosphate AND sn-glycerol 3-phosphate. We have a similar situation for NAD(P)H where we also use two reactions, one for NADH and one for NADPH. So I suggest you treat this as you treat the cases with NAD(P)H.

It would be wrong to change the name to 'glycerol phosphate phosphatase activity' as it then also would include glycerol 2-phosphate.

[pgaudet]

I see 2 options:

1. Keep all 3 terms, add EC:3.1.3.21 on GO:0043136 glycerol-3-phosphatase activity and mention that glycerol-1-phosphate and glycerol-3-phosphate are enantiomers of each other. In this case we may want a grouping term, but that may just be confusing
2. Merge GO:0000121 glycerol-1-phosphatase activity and GO:0043136 glycerol-3-phosphatase activity, and rename something like glycerol-1-phosphatase/glycerol-3-phosphatase activity. They have almost the same number of entries associated; ie maybe all enzymes can catalyze both reactions.

Maybe Option 2 is more consistent with GO practices, but is it confusing ?

Thanks, Pascale

[kaxelsen]

Be careful: glycerol-1-phosphatase = glycerol-3-phosphatase and are both parent to the two reactions sn-glycerol-1-phosphate and sn-glycerol-3-phosphate.

Make sure you write the names correctly as only the 'sn-glyerol's' are unambiguous.

[sjm41]

Thanks both.

Comparing the lists of UniProt IDs annotated with each of the specific RHEAs (1,029 for RHEA:46084 and 1,026 RHEA:66372), I see that they are the same except for 3 entries that only have RHEA:46084 (O33194, P53981, Q9P6N2). So it seems reasonable to think most enzymes are capable of both specific activities.

So I think merging GO:0000121 and GO:0043136 is the correct course of action for GO.

I suggest the final merged term looks like this (with changed aspects shown in bold):

name: glycerol-1-phosphate phosphatase activity def: "Catalysis of the reaction: sn-glycerol-1-phosphate + H2O = glycerol + phosphate, or sn-glycerol-3-phosphate + H2O = glycerol + phosphate". [EC:3.1.3.21] synonym: "alpha-glycerol phosphatase activity" RELATED [EC:3.1.3.21] synonym: "alpha-glycerophosphatase activity" RELATED [EC:3.1.3.21] synonym: "glycerol 3-phosphatase activity" RELATED [EC:3.1.3.21] synonym: "glycerol 3-phosphate phosphohydrolase activity" RELATED [EC:3.1.3.21] synonym: "glycerol-1-phosphate phosphohydrolase activity" RELATED [EC:3.1.3.21] synonym: "glycerol-3-phosphate phosphatase activity" RELATED [EC:3.1.3.21] synonym: "glycerol-3-phosphatase activity" EXACT synonym: "G3PP activity" EXACT xref: EC:3.1.3.21 xref: MetaCyc:GLYCEROL-1-PHOSPHATASE-RXN xref: RHEA:11476 is_a: GO:0016791 ! phosphatase activity

Note that I'm suggesting that we use "glycerol-1-" in the name to address Kristian's point. But I'm also suggesting we use "glycerol-1-phosphate phosphatase activity" rather than "glycerol-1-phosphatase activity" for accuracy. (The two papers I mentioned above use the term 'G3PP', or 'glycerol-3-phosphate phosphatase')