Relationship needed

[gocentral]

Hi,

Please could someone create a relationship between

O:0042346 positive regulation of NF-kappaB import into nucleus and GO:0033160 positive regulation of protein import into nucleus, translocation

and/or their ancestors as appropriate.

Thanks, V

Reported by: vkhodiyar

Original Ticket: geneontology/ontology-requests/8014

[gocentral]

• assigned_to: nobody --> gomidori

Original comment by: mah11

[gocentral]

There are is_a paths from GO:0042346 to GO:0006606 'protein import into nucleus', via 'regulation of protein import into nucleus' (GO:0042306). GO:0033160 refers specifically to regulating the translocation step of protein import into the nucleus (as opposed to docking or release), i.e. it regulates GO:0000060, which is part of GO:0006606.

Because GO:0000060 and GO:0033160 refer to a specific part of the GO:0006606 import process, I don't think we can directly connect GO:0042346 to them -- GO:0042346 could refer to regulating any step of NF-kappaB import. If you have evidence of regulation specifically at the translocation step, we can add terms.

That said, the terms for various proteins aren't all consistent -- we've been careful to make the regulation terms match the regulated processes, but those aren't consistent themselves (e.g. we have NF-kappaB and NFAT import into nucleus, but catenin and SMAD nuclear translocation). Maybe this is an argument for changing over to putting protein IDs in column 16 instead of having the protein-specific terms in the ontology.

m

Original comment by: mah11

[gocentral]

Hi

I had also noticed inconsistancies in this domain, and have opened a SF request to try to find out the difference between GO:0006606 'protein import into nucleus' and GO:0000060 : protein import into nucleus, translocation.

Is it possible to distinguish between protein import into nucleus and protein import into nucleus, translocation if the protein in question is always translocated into the nucleus. I don't think so, any protein involved in the import of SMAD has to be involved in the translocation of SMAD into the nucleus.

Are any transcription factors imported into the nucleus by a process which does not involve translocation? I really don't know.

However, a quick search in google or pubmed with NF-kappa and translocation, suggests that NF-kappaB is translocated into the nucleus, eg PMID: 19169824, and therefore GO:0042348 NF-kappaB import into nucleus should be renamed GO:0042348 NF-kappaB nuclear translocation with the new parent:

GO:0000060 : protein import into nucleus, translocation

The positive regulation relationship terms that Varsha requested therefore make sense. But renaming of these terms is also required.

I agree with Midori that there is not consistancy in this area, would it be possible to find a nuclear import expert (or perhaps someone in GO does know) to find out if any proteins (or transcription factors) are imported into the nucleus by non-translocation methods.

GO:0006607 NLS-bearing substrate import into nucleus probably should be a child of GO:0000060 : protein import into nucleus, translocation as this 'tag' targets the protein to importin.

NFAT also seems to be translocated to the nucleus: PMID: 21088110 so probably this term should be renamed. In addition, the transcription factors which are already stated as being translocated to the nucleus (SMAD, STAT etc) should be child terms of GO:0042991 : transcription factor import into nucleus

Hope this makes sense, bit difficult to see what you are writing in 5 line free text box

Ruth

Original comment by: RLovering

[gocentral]

> I had also noticed inconsistancies in this domain, and have opened a > SF request to try to find out the difference between GO:0006606 > 'protein import into nucleus' and GO:0000060 : protein import into > nucleus, translocation.

The difference between these two terms is clearly represented in the ontology structure:

- protein import into nucleus GO:0006606 ---<p> protein import into nucleus, docking GO:0000059 ---<p> protein import into nucleus, translocation GO:0000060 ---<p> protein import into nucleus, substrate release GO:0000061

GO:0006606 represents the overall process of importing a protein into the nucleus. That import process has three parts, or sub-processes: docking (formation of a complex between the cargo protein, importins, and the nuclear pore), translocation (the actual directional movement of the cargo/importin complex through the pore), and substrate release (dissociation of the cargo from the importins and the pore). GO:0000060 represents the translocation sub-process, and is thus part_of GO:0006606.

I have edited the definitions of all three parts of GO:0006606 to include a bit more detail and explicit genus-differentia info, but the essential distinction between GO:0006606 and GO:0000060 was already represented accurately by the part_of relationships.

> Is it possible to distinguish between protein import into nucleus > and protein import into nucleus, translocation if the protein in > question is always translocated into the nucleus. I don't think so, > any protein involved in the import of SMAD has to be involved in the > translocation of SMAD into the nucleus.

NO. A protein involved in the movement of SMAD into the nucleus might be involved in any of the docking, translocation, or release steps. (Not many particular proteins would be involved all three steps, but not all will be involved exclusively in the translocation part of the import process.)

In fact, I suspect that most of the proteins that are specifically involved in moving, say, SMADs as opposed to any other proteins, into the nucleus are more likely to act in the docking part of import, because that's when cargo and carrier proteins (aka "receptor" proteins, often importin alpha/beta heterodimers) get together.

I think the '[x] protein nuclear translocation' terms should be renamed 'x protein import into nucleus', moved up to be direct is_a children of GO:0006606 (or GO:0006607, where enough is known), and have definitions reworded accordingly.

> Are any transcription factors imported into the nucleus by a process > which does not involve translocation? I really don't know.

> However, a quick search in google or pubmed with NF-kappa and > translocation, suggests that NF-kappaB is translocated into the > nucleus, eg PMID: 19169824, and therefore GO:0042348 NF-kappaB > import into nucleus should be renamed GO:0042348 NF-kappaB nuclear > translocation with the new parent:

> GO:0000060 : protein import into nucleus, translocation

No, as noted above, I would prefer to rename the 'translocation' terms.

Your search results indicate that the distinction between import and the translocation part of import is not always made clear in literature that focuses on the biological roles of specific proteins and the circumstances under which they enter the nucleus. In other words, a lot of papers on SMADs, NF-kappaB, etc. use "translocation" in a broader sense than GO:0000060 does. The translocation-part-of-import distinction is made in literature describing the import mechanism. It's unfortunate that usage in the literature isn't more consistent overall, but GO has to deal with exactly this sort of situation all the time.

It's because GO has to impose consistency throughout the ontology -- even if such consistency isn't found in the literature -- that I want to rename the 'translocation' terms in question. I also think that's by far the safer approach for any existing annotations.

> The positive regulation relationship terms that Varsha requested > therefore make sense. But renaming of these terms is also required.

No, the relationship that Varsha requested cannot be added, because it would put a whole process (broader with respect to the import aspect; narrower only in that it refers to a specific class of protein) under a part of the overall process. This would be a TPV.

I've noted which terms should be renamed above.

> I agree with Midori that there is not consistancy in this area,

The inconsistency I spotted is that most of the protein-specific terms in this branch of the ontology are named and placed (wrongly) as translocation terms, whereas the terms for NFAT and NF-kappaB are named and placed as import terms. I've noted above that I think the latter is ontologically correct, and I would like to rename the 'translocation' terms to 'import'.

> GO:0006607 NLS-bearing substrate import into nucleus probably should > be a child of GO:0000060 : protein import into nucleus, > translocation as this 'tag' targets the protein to importin.

No, because GO:0006607 refers to the entire import process, not just the translocation step.

> NFAT also seems to be translocated to the nucleus: PMID: 21088110 so > probably this term should be renamed.

Actually, this and the NF-kappaB term are the ones that don't need to be renamed.

> In addition, the transcription factors which are already stated as > being translocated to the nucleus (SMAD, STAT etc) should be child > terms of GO:0042991 : transcription factor import into nucleus

OK

> Hope this makes sense, bit difficult to see what you are writing in > 5 line free text box

I usually type elsewhere and then copy/paste into the SF comment box.

m

Original comment by: mah11

[gocentral]

Hi Midori

sorry many of my suggestions became redundant once you made the excellent decision: the '[x] protein nuclear translocation' terms should be renamed 'x protein import into nucleus', moved up to be direct is_a children of GO:0006606 (or GO:0006607, where enough is known), and have definitions reworded accordingly.

I completely agree with this suggestion and now appreciate the 3 stages of protein import. However, the presence of these 3 children to a non-expert does not necessarily mean that proteins are only imported into the nucleus by translocation. If proteins can get into the nucleus by non-translocation means then possibly these 3 terms would still exist.I also find the position of GO:0006607 NLS-bearing substrate import into nucleus hard to understand, which I think, based on the docking definition should be a parent of the docking, translocation and release terms, if all three terms are intimately related as described.

Thank you for agreeing to clarify the definitions

Ruth

Original comment by: RLovering

[gocentral]

Hi Ruth,

I'm glad the suggestion to rename those 'translocation' terms makes sense! I'll get on to it later this week.

I don't think there is a mechanism for nuclear protein import that doesn't involve translocation. As far as I can tell, there may be small proteins that can diffuse through the nuclear pore, but most (even some that should be small enough) don't, and I'm only finding the one model of docking/translocation/release for actively transporting proteins through the pore. There seem to be some minor variations in specifics, often where there isn't enough data to distinguish between possible mechanistic minutiae, but no big differences.

I'm not sure why GO:0006607 was included in the first place -- it dates back to the very earliest days of GO -- nor am I sure how broadly/narrowly we should interpret 'NLS'. Of proteins that go into the nucleus, some have 'classical' basic NLSs, some have other identified signals, and for others we don't know; in some cases proteins without NLSs may be able to get into the nucleus by associating with NLS-bearing proteins.

I don't want to move the parts of GO:0006606 (GO:0000059,60,61) down to part_of GO:0006607, because the same steps take place whether the cargo protein has an identified NLS or not. The major difference is apparently which carriers (importins and the like) associate with the cargo. Maybe we should merge GO:0006607 into GO:0006606; the child doesn't have nearly as many annotations as the parent.

m

Original comment by: mah11

[gocentral]

p.s. some reviews I've found useful:

PMID:9126736 http://www.nature.com/nature/journal/v386/n6627/abs/386779a0.html PMID:14570049 http://www.nature.com/nrm/journal/v4/n10/full/nrm1230.html PMID:14504656 http://www.springerlink.com/content/bxfwkda38ure1quk/ PMID:18282135 http://protein.bio.msu.ru/biokhimiya/contents/v72/full/72131439.html

Original comment by: mah11

[gocentral]

I've now renamed and moved the terms that had 'translocation' in the names but didn't really refer only to the translocation step of nuclear protein import. That should help make it clearer that the existing relationships for GO:0042346 are correct and complete.

m

Original comment by: mah11

[gocentral]

• status: open --> closed-fixed

Original comment by: mah11