geneontology / go-ontology

Source ontology files for the Gene Ontology
http://geneontology.org/page/download-ontology
Creative Commons Attribution 4.0 International
220 stars 40 forks source link

tRNA ligase activity, involved in tRNA splicing #3037

Closed gocentral closed 9 years ago

gocentral commented 18 years ago

Hi,

This item has been taken out of another sourceforge entry 'TAIR function terms-nov 2005'. I have included all the discussion on this subject from the previous entry.

Requested by Suparna:

6(Better name) tRNA ligase activity involved in

tRNA splicing ISA tRNA ligase activity GO:0004812 Def: three intrinsic activities: an N-terminal adenylyltransferase domain that resembles T4 RNA ligase 1, a central domain that resembles T4 polynucleotide kinase (without 30-phosphatase activity) and a C-terminal cyclic phosphodiesterase (CPDase) domain that is reminiscent of the 2H phosphotransferase enzyme superfamily (13–15). The multifunctional yeast tRNA ligase performs a series of reactions: (i) the 20,30-cyclic phosphate at the end of the 50 tRNA half is hydrolysed by the CPDase activity to generate a 20-phosphate and 30-OH; (ii) the 50-OH group at the end of the 30 tRNA half is phosphorylated by the GTP-dependent kinase activity and (iii) the ligase protein is adenylylated and the AMP is transferred to the 50-phosphate of the 30 tRNA half followed by the formation of the 20- phosphomonoester–30,50-phosphodiester bond and the release of AMP Ref: PMID: 15653639

Date: 2005-11-30 13:39 Sender: kchris Logged In: YES user_id=473890

Hi Suparna,

For your #6, the term you're talking about "tRNA ligase activity ; GO:0004812" is NOT the enzyme that is involved in splicing the introns that occur in tRNAs (and occasionally in other RNA types). This term refers to the class of activities that charge a tRNA by adding an amino acid to the tRNA. Currently this term has two roots to parent:

molecular_function .catalytic activity ..ligase activity ...ligase activity, forming phosphoric ester bonds ....RNA ligase activity .....tRNA ligase activity ...ligase activity, forming carbon-oxygen bonds ....ligase activity, forming aminoacyl-tRNA and related compounds .....tRNA ligase activity

It might be worth checking whether this type of tRNA ligase should really have parentage under "RNA ligase activity" and thus under "ligase activity, forming phosphoric ester bonds". I kind of wonder if it should really have parentage under both that term AND "ligase activity, forming carbon-oxygen bonds", but the review I'm looking at is only on the splicing of introns from tRNAs, not the charging reaction, so I'll leave that for the moment, and hopefully someone else to check out. It might also be worth investigating if there is another name that could be used for this term that is less ambiguous and the "tRNA ligase" name could be made a synonym.

Going back to the subject of splicing introns from tRNA, the tRNA ligase that performs this in S. cerevisiae is a multifunctional enzyme with three distinct activities: a phosphodiesterase activity, a polynucleotide kinase activity, and RNA ligase activity. It may be worth noting that metazoans seem to have two different redundant mechanisms for religating the two halfs of a cleaved pre- tRNA, one of which is mechanistically the same as that of cerevisiae and one which is not. The mechanism is not yet known in Archaea.

in SGD currently, the TRL1 gene, which encodes the tRNA ligase that is involved in splicing is annotated to only one function term: "RNA ligase (ATP) activity ; GO:0003972". Clearly, it should be annotated to additional function terms. I am not yet sure what terms should be used, or whether appropriate terms already exist. I'll look into that in the next week or so because SGD should really have this enzyme fully annotated.

-Karen

Date: 2005-11-23 04:27 Sender: gomidori Logged In: YES user_id=436423

> 6. (Better name) tRNA ligase activity involved in > tRNA splicing

I agree that this needs a different name - it's not consistent with our function term guidelines to make distinctions, or name terms, in the function ontology based on which process an activity or gene product is involved in. Also, it sounds like the protein you want to annotate has several activities -- but presumably Karen will find out what's going on.

Reported by: smundodi

Original Ticket: "geneontology/ontology-requests/3047":https://sourceforge.net/p/geneontology/ontology-requests/3047

gocentral commented 18 years ago

Logged In: YES user_id=546388

The term of ligase for the aminoacyl-tRNA synthetases dates back quite a while; a ligase refers to an enzyme requiring ATP that has generates an intermediate of AMP-x, where X can be an amino acid (as in the case of the aminoacyl-tRNA synthetase (amino-acid adenylate), or, as in a (DNA/RNA) ligase , AMP-P-5'-Nucleic acid. Some definitions of a ligase actually say that they also called synthetase, but I have only seen the term "syntheteae" used in reference ot the aminoacyl-tRNA synthetase. GO:0004812 doesn't even have synthetase as a synonym. EC uses aminoacid-tRNA ligase as an official name, so I would definitely suggest that the main name be changed to aminoacyl-tRNA ligase activity, with aminoacyl-tRNA sythetase as a synonym. While I'm on it, I notice that glutamate -trNA (Gln) ligase (GO:0050561), as defined should be a child of 4812, not a sib. also GO:0050562, as defined is also a child. also GO:0050560, as defined is also a chiid, not a sib.

Original comment by: hdrabkin

gocentral commented 18 years ago

Logged In: YES user_id=473890

Hi Harold,

Thanks for comments on a more appropriate name for GO:0004812. That sounds good to me.

The other big question that I had was the current dual parentage of this term.

molecular_function .catalytic activity ..ligase activity ...ligase activity, forming phosphoric ester bonds ....RNA ligase activity .....tRNA ligase activity ...ligase activity, forming carbon-oxygen bonds ....ligase activity, forming aminoacyl-tRNA and related compounds .....tRNA ligase activity

I agree with the parentage that goes up through "ligase activity, forming carbon-oxygen bonds". However, I am quite skeptical of the parentage going up through "RNA ligase activity" and its parent term "ligase activity, forming phosphoric ester bonds".

Looking at the listings for the EC groupings 6.1.1 and 6.5.1 is consistent with my feeling that the parentage under "RNA ligase" activity is incorrect for the term "tRNA ligase ; GO:0004812".

EC 6.1.1 - Ligases Forming Aminoacyl-tRNA and Related Compounds Contents

EC 6.1.1.1 tyrosine—tRNA ligase EC 6.1.1.2 tryptophan—tRNA ligase EC 6.1.1.3 threonine—tRNA ligase EC 6.1.1.4 leucine—tRNA ligase EC 6.1.1.5 isoleucine—tRNA ligase EC 6.1.1.6 lysine—tRNA ligase EC 6.1.1.7 alanine—tRNA ligase EC 6.1.1.8 deleted EC 6.1.1.9 valine—tRNA ligase EC 6.1.1.10 methionine—tRNA ligase EC 6.1.1.11 serine—tRNA ligase EC 6.1.1.12 aspartate—tRNA ligase EC 6.1.1.13 D-alanine—poly(phosphoribitol) ligase EC 6.1.1.14 glycine—tRNA ligase EC 6.1.1.15 proline—tRNA ligase EC 6.1.1.16 cysteine—tRNA ligase EC 6.1.1.17 glutamate—tRNA ligase EC 6.1.1.18 glutamine—tRNA ligase EC 6.1.1.19 arginine—tRNA ligase EC 6.1.1.20 phenylalanine—tRNA ligase EC 6.1.1.21 histidine—tRNA ligase EC 6.1.1.22 asparagine—tRNA ligase EC 6.1.1.23 aspartate—tRNAAsn ligase EC 6.1.1.24 glutamate—tRNAGln ligase EC 6.1.1.25 lysine—tRNAPyl ligase

EC 6.5.1 Contents

EC 6.5.1.1 DNA ligase (ATP) EC 6.5.1.2 DNA ligase (NAD+) EC 6.5.1.3 RNA ligase (ATP) EC 6.5.1.4 RNA-3'-phosphate cyclase

-Karen

Original comment by: krchristie

gocentral commented 18 years ago

Logged In: YES user_id=546388

IF ..ligase activity forming phosphoric ester bonds, THEN ... ultimate bond formation of the reaction catalyzed is NOT a phosphoric ester for aminoacyl-tRNA ligase. Thus, the parentage seems wrong...but..

The reaction actually occurs in two steps:

  1. Amino acid + ATP == aminoacid-AMP + PP (one can actually measure this)

  2. aminoacid-AMP + tRNA == aminoacid-tRNA + AMP

See for example Fersht, Enzyme Structure and mechanism, 2nd ed.) So, there IS a phosphoric acid ester in the reaction mechanism. This might have been what was meant.

Original comment by: hdrabkin

gocentral commented 18 years ago

Logged In: YES user_id=473890

I can see your point, that it does at some point make a phosphoric ester bond. However, I also wonder if the additional parentage was added for a reason not involving such sophisticated understanding of the reaction, but rather on the basis of lack of understanding of what the term represented. As it is clear that it is easy to misinterpret the existing term as being involved in RNA ligation as its main purpose, I wonder if the parentage was added for that reason.

The existing representation of the 6.1.1 enzymes under the 6.5.1 branch seems inconsistent with the EC nomenclature, so if we keep the RNA ligase (6.5.1.-) parentage, the definition should be expanded to make it clear what is going on. Alternately, maybe we should only be representing the final bond, not an intermediate mechanism. I quess the decision comes down to which course would be most consistent with GO practice. It seems to me that if you feel that the tRNA synthetases are performing two reactions, GO would typically represent those with two separate terms. If it's merely a reaction intermediate, then I'm not sure if we need to represent it. My 2p at the moment.

-Karen

Original comment by: krchristie

gocentral commented 18 years ago

Logged In: YES user_id=546388

I'm glad you brought that up, because I was going to include a more general question about whether we should only reflect the final products, and not the intermediates. However, AMP-AA is stable enough to be isolated I think.

I'm sort of in favor of removing the phosphoric acid ester parentage myself; 2nd best is to make it clear in the definition what is going on. But I wouldn't want this split into two reactions.. since if you had rxn 2, you would have to have had rxn 1.

Original comment by: hdrabkin

gocentral commented 18 years ago

Logged In: YES user_id=473890

Personally, I think we really only need to represent the final bond produced, so I'm in agreement that we should remove the phosphoric acid ester parentage.

Original comment by: krchristie

gocentral commented 18 years ago

Logged In: YES user_id=546388

Hi midori

If one would ask what does an aminoacyl-tRNA synthetase/ligase DO, it would be AA + tRNA + ATP ==AA-tRNA

I don't think we want to make separate terms for HOW it does it (2-step process). If we go that route, we would need to do it for a great deal of reactions that use co-factors. In this case, it would even include every single ligase (DNA, RNA, etc., since they always involve 3 reactants: donor, acceptor, and ATP); unless a reaction is concerted, there will always be a step 1 then step 2 in these reactions.

Original comment by: hdrabkin

gocentral commented 18 years ago

Logged In: YES user_id=436423

Keeping it as one reaction/one function term is fine from the GO practice perspective -- there's plenty of precedent for counting 'coupled' reactions as one molecular function.

m

Original comment by: mah11

gocentral commented 18 years ago

Logged In: YES user_id=546388

I think then we should make the name "aminoacyl-tRNA ligase (not tRNA ligase), move the sibs to children of it; This leaves the dual parentage. If the current dual parentage appears confusing, we may want to remove it. Since one of them is refering to an intermedate step, then we might want to think about this in the future. If the parents are to cover intermediate steps this could get messy... at present, if one considers all of the "subfunctions" that go into making aminoacyl-tRNA ligase activity, they would be tRNA binding amino acid binding ATP binding, and the catalysis of the two reactions ATP+AA ==AMP-AA + PP AMP-AA + tRNA == AA-tRNA + AMP

Original comment by: hdrabkin

gocentral commented 18 years ago

Logged In: YES user_id=473890

Hi,

Let me restate, just to make sure I haven't misunderstood your proposal.

  1. Rename the term "tRNA ligase activity ; GO:0004812" to "aminoacyl-tRNA ligase activity".

  2. Move these 4 terms, which are currently siblings of "tRNA ligase activity ; GO:0004812" to be children of this term instead.

aspartate-tRNA(Asn) ligase activity ; GO:0050560 def: "Catalysis of the reaction: tRNA(Asx) + L-aspartate + ATP = aspartyl- tRNA(Asx) + pyrophosphate + AMP." [EC:6.1.1.23, MetaCyc:6.1.1.23-RXN] exact_synonym: "aspartate-tRNAAsn ligase activity" []

D-alanine-poly(phosphoribitol) ligase activity ; GO:0047473 def: "Catalysis of the reaction: poly(ribitol phosphate) + D-alanine + ATP = O-D-alanyl-poly(ribitol phosphate) + pyrophosphate + AMP." [EC:6.1.1.13, MetaCyc:6.1.1.13-RXN] subset: gosubset_prok

glutamate-tRNA(Gln) ligase activity ; GO:0050561 def: "Catalysis of the reaction: tRNA(Glx) + L-glutamate + ATP = glutamyl- tRNA(Glx) + pyrophosphate + AMP." [EC:6.1.1.24, MetaCyc:6.1.1.24-RXN] exact_synonym: "glutamate-tRNAGln ligase activity" []

lysine-tRNA(Pyl) ligase activity ; GO:0050562 def: "Catalysis of the reaction: tRNA(Pyl) + L-lysine + ATP = L-lysyl-tRNA(Pyl)

  1. Remove "RNA ligase" parentage of GO:0004812 since it is relevant to an intermediate step in the coupled reaction, not to the final bond produced.

This all sounds good to me. If I misunderstood anything, please let me know.

-Karen

Original comment by: krchristie

gocentral commented 18 years ago

Logged In: YES user_id=546388

Hi Karen

This one, however, should not be moved:

D-alanine-poly(phosphoribitol) ligase activity ; GO:0047473 def: "Catalysis of the reaction: poly(ribitol phosphate) + D-alanine + ATP = O-D-alanyl-poly(ribitol phosphate) + pyrophosphate + AMP." [EC:6.1.1.13, MetaCyc:6.1.1.13-RXN] subset: gosubset_prok

There is no tRNA in this reaction. The only thing it has in commone is an aminoacyl-adenylate, I think.

Original comment by: hdrabkin

gocentral commented 18 years ago

Logged In: YES user_id=473890

OK, makes sense. It seems to me like this aspect of the question, dealing with the current term 'tRNA ligase activity ; GO:0004812' is essentially resolved.

Related to this area of the ontology, I was just noticing that the parent term of 'tRNA ligase', the term 'ligase activity, forming aminoacyl-tRNA and related compounds ; GO:0016876' does not have a definition. Do you know enough to suggest one?

Regarding the other aspect of the initial question, about the tRNA ligase involved in RNA splicing, Jodi has volunteered to look into the annotations for SGD's TRL1 gene. So, we should know in the next week or so if new term requests are needed, though if you want to resolve this item and close it, we'd be perfectly happy to just submit NTRs specifically for any needed terms.

-Karen

Original comment by: krchristie

gocentral commented 18 years ago

Logged In: YES user_id=473890

Hi, especially to Harold,

Have we resolved on what changes to make? I'm typing from memory because I need to leave work, but I think we need to:

  1. change term name as Harold suggested
  2. remove some parentage of the term, so that it does not get parentage under the 'RNA ligase activity' branch
  3. We could add a bit to the definition so that it is obvious to non-experts of tRNA metabolism that the reaction described is that of charging tRNAs with the appropriate amino acid.

Harold, are you up for making these changes this week? If not, I'd be happy to give it a go, but we just need to spell it clearly what changes are appropriate.

thanks,

-Karen

Original comment by: krchristie

gocentral commented 18 years ago

Logged In: YES user_id=546388

Just saw this buried in the text. Ok, let me do this sometime this afternoon (thursday)....

Original comment by: hdrabkin

gocentral commented 18 years ago

Logged In: YES user_id=546388

Ok I just

  1. renamed tRNA ligase activity to aminoacyl-tRNA ligase activity.
  2. added synonym of aminoacyl-tRNA synthetase activity
  3. added comment describing the bond made
  4. removed RNA ligase activity as a parent
  5. moved 3 sibling terms of "XX-tRNA (X) ligase activity to children of aminoactyl-tRNA ligase activty

I notice that RNA ligase has no def; but the RNA ligase (ATP) activity has a def that I would think is the same as RNA ligase activity. The 2'5' RNA ligase activity would still use ATP. Thus, I would suggest

  1. obsolete RNA ligase activity, GO:0008452, but move it's synonym to GO:0003972
  2. Remove (ATP) from the term name of GO:0003972
  3. make 2'5 RNA ligase activity a child of GO:0003972.
  4. make new term 3'5' RNA ligase activity as a child of GO:0003972 also. This would thus distinquish the two types of RNA ligase.

Original comment by: hdrabkin

gocentral commented 18 years ago

Logged In: YES user_id=546388

Did I say oboslete!!???? No, actually, since it has no def, we can just merge it into RNA ligase (ATP) activity.

Original comment by: hdrabkin

gocentral commented 18 years ago

Logged In: YES user_id=473890

Hi Harold,

Thanks for working on the 'tRNA ligase activity' term. That should really clear things up.

As for your suggestions on RNA ligase, I am uncertain about this. Not knowing much about RNA ligases, I went to check into it a little. Not that I feel that I have a good handle on it yet, but from what I have seen so far, I am not so sure that 2'-5' RNA ligases "would still use ATP". Do you have good info on the mechanism of 2'-5' RNA ligases?

Here's some of what I came across that makes me wonder if "2'-5' RNA ligases" should be under "RNA ligase (ATP) activity".

This paper: Shuman S, Lima CD. The polynucleotide ligase and RNA capping enzyme superfamily of covalent nucleotidyltransferases. Curr Opin Struct Biol. 2004 Dec;14(6):757-64. PMID: 15582400

says: ATP-dependent RNA ligases join 3′-OH and 5′-PO4 RNA termini through the same series of three nucleotidyl transfer steps involving activated covalent enzyme–adenylate and RNA–adenylate intermediates (Figure 1). There are two branches of the RNA ligase family, exemplified by bacteriophage T4 RNA ligase 1 (Rnl1) and RNA ligase 2 (Rnl2) 4. and 5.. Rnl2-like proteins are found in all three phylogenetic domains. The Rnl2 family includes the mRNA-editing ligases of Trypanosoma and Leishmania [6], plus RNA ligases encoded by vibriophage, mycobacteriophage, eukaryotic viruses and many species of archaea [4]. Rnl1-like ligases are few in number and have a relatively narrow phylogenetic distribution, which is limited, as far as we know, to bacteriophages, fungi and baculoviruses 7. and 8..   This paper: Daniel R. Semlow, Scott K. Silverman (2005) Parallel Selections In Vitro Reveal a Preference for 2¢–5¢ RNA Ligation upon Deoxyribozyme-Mediated Opening of a 2',3'-Cyclic Phosphate. J Mol Evol (2005) 61:207–215

talks about "native 3'-5' linkages" and "non-native 2'-5' linkages"

-Karen

Original comment by: krchristie

gocentral commented 18 years ago

Logged In: YES user_id=546388

Actually, I just remembered that if we are refering to the tRNA splicing, ATp IS involved, but but in a different way

  1. endonuclease cleaves to leave a 2-3 cylic phosphate
  2. cylic 2'3' is opened at the 3' phosphate by 2'3'[ cyclic phosphodieseterase; the 5'j)H of the 3'1/2 of the tRNA is phosphorylated by a polynucleotide kinase activity. Then the protein is adenylated and transfers its AMP residue to the 5' phosphate end of the 3' half of the tRNA, and the ligase then joins the twi halves.
  3. an NAD-dependent 2' phosphtransferase removes the 2'PO4 at the splice junction. (most of this has been worked out in yeast; Review by Westaway and Abelson in ISBN:155581073X)

SO, based on the fact that the 'tRNA ligase activity" has a synonym that refers to the tRNA ligase enzyme (involved in the UPR pathways), forget my proposal for the merge. However, we clearly need a proper term name and defintion.

I think at the least, it should be renamed. "tRNA ligase activity involved in tRNA splicing"

Original comment by: hdrabkin

gocentral commented 17 years ago

Original comment by: mah11

gocentral commented 13 years ago

Original comment by: mah11