Missing parent: GO:0140818 | mRNA 5'-phosphatase activity

[ValWood]

• GO term ID and label for which you request a new superclass

GO:0140818 | mRNA 5'-phosphatase activity does not have superclass

GO:0004651 | polynucleotide 5'-phosphatase activity

[pgaudet]

Something is not right about GO:0004651 | polynucleotide 5'-phosphatase activity -

GO ID	Label	EC	RHEA	EC & RHEA Description
GO:0140818	mRNA 5'-phosphatase activity	EC:3.6.1.74	RHEA:67004
GO:0004651	polynucleotide 5'-phosphatase activity	EC:3.1.3.3	RHEA:11008	purine ribonucleoside in mRNA

• EC:3.1.3.33 and RHEA:11008 has a single entry in UniProt: A0A7H0DN85, annotated to EC:3.1.3.33 by similarity
• RHEA:11008 has a note that the more general form of the reaction id RHEA:67004
• RHEA is citing PMID:21956912 which may be useful

@raymond91125 can you please investigate this? We probably need to obsolete GO:0004651 as it has been may have been used to mean GO:0140818

Thanks, Pascale

[pgaudet]

Related to #24892

[raymond91125]

The term relationship should be: GO:0140818 "mRNA 5'-phosphatase activity" EC 3.6.1.74 covers both purine and pyrimidine substrates ... GO:0004651 "polynucleotide 5'-phosphatase activity" EC 3.1.3.33 covers ONLY PURINE substrates Both are involved in mRNA 5' capping.

[sjm41]

Another odd thing here is the EC number (EC:3.1.3.33) for GO:0004651 polynucleotide 5'-phosphatase activity: if this activity is removing one phosphate from "a 5'-end triphospho-(purine-ribonucleoside) in mRNA", then I think it should have a EC:3.6 (Hydrolases - Acting on acid anhydrides) prefix rather than 3.1 (Hydrolases - Acting on ester bonds). Then the new relationship suggested by @raymond91125 would make sense.

What do you think about the EC:3.1.3.33 number here @kaxelsen ?

[kaxelsen]

I think you are right, especially since the comment says: "Does not act on nucleoside monophosphates", which is where the ester bond is. I will look into it and suggest a transfer of the EC number to the subclass EC 3.6

[kaxelsen]

@sjm41 I'm sorry, but my first impression of the problem was wrong. After reading the evidence for the EC 3.1.3.33 entry (PMID: 4289819 ), I have learned that the entry only covers 5'-MONOphosphorylated polynucleotides, so it is the Rhea mapping that is completely wrong. So in the table above it is the Rhea mapping and the EC & Rhea description that is wrong. The substrate of EC 3.1.3.33. is a "a 5'-end phospho-(nucleoside) in RNA or DNA"

We are unfortunately in the middle of a freeze, so I will not be able to go further with this problem for the next couple of days, and I'm not sure I can create a Rhea reaction as the enzyme works on both RNA and DNA.

[sjm41]

Thanks @kaxelsen

It sounds like the GO:0004651 definition (which is attributed to EC:3.1.3.33) is actually correct then!: "Catalysis of the reaction: 5'-phosphopolynucleotide + H2O = polynucleotide + phosphate." [EC:3.1.3.33] (and we should ignore the current (incorrect) EC def that says it's specific for purines)

Which means the action item for GO here is just to remove the RHEA:11008 xref from GO:0004651? (on the understanding that the EC:3.1.3.33 reaction definition will get corrected in time)

Though maybe we should also change the GO term name to "polynucleotide 5'-monophosphatase activity" to be clearer?

==

id: GO:0004651 name: polynucleotide 5'-phosphatase activity namespace: molecular_function def: "Catalysis of the reaction: 5'-phosphopolynucleotide + H2O = polynucleotide + phosphate." [EC:3.1.3.33] synonym: "5'-polynucleotidase activity" RELATED [EC:3.1.3.33] synonym: "polynucleotide 5'-phosphohydrolase activity" RELATED [EC:3.1.3.33] synonym: "polynucleotide 5'-triphosphatase activity" RELATED [EC:3.1.3.33] xref: EC:3.1.3.33 xref: MetaCyc:RXN-21861 xref: RHEA:11008 is_a: GO:0016791 ! phosphatase activity property_value: term_tracker_item "https://github.com/geneontology/go-ontology/issues/25810" xsd:anyURI

[kaxelsen]

For your information: EC 3.1.3.33 is now linked to RHEA:67772 and RHEA:67776, both public

[raymond91125]

For your information: EC 3.1.3.33 is now linked to RHEA:67772 and RHEA:67776, both public

I'm not seeing the reference to EC 3.1.3.33 on pages of RHEA:67772 and RHEA:67776. The RHEAs do refer to MetaCyc RXN-21861 which then refers to EC 3.1.3.33. But information about the EC is different at the three different sites: Brenda, Expasy, and enzyme-database.

For https://amigo.geneontology.org/amigo/term/GO:0004651, the text definition is presently different from the participants. As @sjm41 mentioned, should we wait for EC to be corrected? That is the only XREF for GO:0004651. Also, GO:000465 EXP annotations appear to be incompatible with the definition found in PMID:4289819, RHEA:67772, or RHEA:67776.

[kaxelsen]

@raymond91125 The changes have been made in the working copy of ENZYME. They will unfortunately only become public in UniProt release 2024_02 (that includes releases of ENZYME and Rhea).

[ValWood]

I haven't fully followed the changes here, but was O:0004651 name: polynucleotide 5'-phosphatase activity correct for the mRNA capping enzyme (i.e will it be a parent of mRNA 5'-triphosphate monophosphatase activity) or do I need to get these annotations/PAINT mappings updated?

If so can do this in advance of the GO/Rhea changes.

[raymond91125]

I was hoping to see the new release of UniProt to make sure exactly what we are going to do. Just a couple of days more?

[ValWood]

Ok, not in a hurry, I just need to remember not to forget that I need to do something ;)

[raymond91125]

For your information: EC 3.1.3.33 is now linked to RHEA:67772 and RHEA:67776, both public

@kaxelsen I'm not seeing this in Release of 27-Mar-24, https://enzyme.expasy.org/index.html. Am I missing something? Thanks.

[raymond91125]

To me, this is all still very confusing. EC and RHEA have not been updated (publicly) as described. As it stands, GO:0004651 polynucleotide 5'-phosphatase activity, the only DEF XREF is EC:3.1.3.33 which is purine ribonucleoside specific (at https://enzyme.expasy.org and RHEA), but different at https://www.enzyme-database.org and MetaCyc.

Parenthetically, PMID:4289819 described 3 different activities, one of those is a RNA/DNA 5' phosphatase. "A third unique phosphatase activity was also encountered in extracts of phage-infected cells. An enzyme, purified extensively from this source, was found capable of removing the 5’-phosphate groups of ribo- and deoxyribopolynucleotides", and "The final preparation shows a high degree of specificity in the cleavage of 5’-phosphoryl groups from polynucleotides and is without effect on 5’-mononucleotides."

[raymond91125]

@ValWood I think the answer to your original question is that GO:0004651 | polynucleotide 5'-phosphatase activity is probably not a superclass of GO:0140818 | mRNA 5'-phosphatase activity

[raymond91125]

I changed the DEF and XREF to that of RHEA:11008. This is neither perfect nor final but hopefully EC can get their fixes in before we do anything more. Please let me know if anyone thinks more should be done now. Thanks.

[kaxelsen]

@raymond91125 The changes have been made in the working copy of ENZYME. They will unfortunately only become public in UniProt release 2024_02 (that includes releases of ENZYME and Rhea).

I'm sorry. The entries did not become public in release 2024_02 as I expected. But release 2024_03 is May 29 so it should all be good by then.

[ValWood]

OK, I will report the mapping for

GO:0004651 | polynucleotide 5'-phosphatase activity | IBA with CAL0000199068 , PTN001999388 , S000006149 , S000004792

[ValWood]

I'm still a bit confused what should be annotated to GO:0004651 | polynucleotide 5'-phosphatase activity...

[ValWood]

RIght in that case the existing annotations are correct. But the enzyme is not specific for purine (as far as I know) so why do we need this term?

[raymond91125]

Let's wait until the end of May to see what needs to be done. Thanks.

[raymond91125]

@ValWood Given the revised EC/RHEA, GO:0004651 is not a superclass of GO:0140818.

[raymond91125]

Please reopen if more is to be done.