ntrs and ontology changes: ER to Golgi transport related

[gocentral]

We've done an extensive review of the cerevisiae COPII genes involved in secretion from the ER to the Golgi and would like to ask for some new terms and some ontology changes.

1) Could we merge GO:0048221 "rough ER to cis-Golgi vesicle-mediated transport" into GO:0006888 "ER to Golgi vesicle-mediated with GO:0006888 being the primary? Based on comparing the definitions/children/references, these seem to be the same process.

2) Add the following new terms:

ER vesicle budding GO:new1 def: The evagination of the ER membrane, resulting in formation of a vesicle synonyms: ER-derived vesicle budding source: PMID:10219233 relationships: membrane budding GO:6900 .is_a ER vesicle budding GO:new1 ER too Golgi vesicle-mediated transport .part_of ER vesicle budding GO:new1

ER transport vesicle coat formation GO:new2 def: The addition of specific coat proteins to ER membranes during the formation of transport vesicles. synonyms: COPII coat formation (narrow?) source: PMID:10219233 relationships: vesicle coating GO:6901 .is_a ER transport vesicle coat formation
ER vesicle budding GO:new1 .part_of ER transport vesicle coat formation (as a side note, we didn't like the name "ER transport vesicle coating" because it sounded like a component to us. Can we change the rest of the "coating" term names to coat formation?)

regulation of ER transport vesicle coat formation GO:new3 def: standard relationships: standard under GO:new2 and GO:60628

3) Please obsolete GO:0048208 'COPII coating of Golgi vesicle'. This term/definition do not represent the biology and are not supported by the given pubmed source. COPII vesicles and coat proteins assemble on the ER membranes. No one has used this term yet and you could put "please consider using" with GO:new2 or GO:48200.

Alright, I think that's it. Feel free to ask if anything is unclear or just incorrect.

Thanks! -Julie

Reported by: juliep

Original Ticket: geneontology/ontology-requests/6781

[gocentral]

assigning to D&T because they've been working on this area at the ASCB meeting ... good timing!

Original comment by: mah11

[gocentral]

• assigned_to: nobody --> tairtb

Original comment by: mah11

[gocentral]

Hi Julie,

We heard some fantastic talks at ASCB about this and have already started to work on this process. We will look over this SF item and let you know if we have questions. Please have a look at what we've already done. I think a lot of it already addresses your concerns.

David (& Tanya who is currently on a plane back home)

Original comment by: ukemi

[gocentral]

Hi David (and Tanya),

I'll gladly take a gander at what you've already done...but where am I supposed to look?

Thanks! -Julie

Original comment by: juliep

[gocentral]

Hi Julie,

In case David & Tanya don't have a chance to respond right away -- they've added new terms with IDs GO:0090110-90114 and 90164-90168 , and mostly under 'ER to Golgi vesicle-mediated transport' GO:0006888 and 'Golgi organization' GO:007030. This is all now live.

cheers, m

Original comment by: mah11

[gocentral]

Hi Julie,

You can also search on the string 'COPII' to see what else we put in. Liz Miller from Columbia University spoke at the meeting and I hope that we were able to capture what she shared accurately.

If we missed stuff (or got it wrong), please let us know and we'll be happy to add/edit it.

Cheers,

Tanya

Original comment by: tberardini

[gocentral]

Thanks Tanya and David,

So we (Eurie, Rama, and I) had a few questions and some additional things we'd like added/edited.

Edits: 1) Could the CC term 'COPII vesicle coat' GO:30127 get some sort of parentage up to ER? 2) Please add the following synonyms for GO:90114 'COPII vesicle budding' -ER exit (narrow) -COPII vesicle formation (exact) -ER vesicle budding (exact? broad?) 3) Please add the following synonyms for GO:48208 'COPII coating of ER-derived vesicle' -COPII vesicle coat formation (exact) -ER vesicle coat formation (exact? broad?) 4) Merge GO:48221 'rough ER to cis-Golgi vesicle-mediated transport' into GO:6888 'ER to Golgi vesicle-mediated transport'. These two terms seem to be redundant. 5) We would like a new term 'regulation of COPII vesicle coating' synonym: 'regulation of COPII coating of ER-derived vesicle' (exact) standard defs and parentage. We're not sure if it can be a direct child of GO:60628 or whether you need to create a new term 'regulation of COPII vesicle budding' to be a parent of this term, but we leave it up to you.

Questions: 1) Other than Liz Miller, were there any other speakers who talked on secretion for whom we can look up their papers? 2) For what exactly was GO:90113 'regulation of ER to Golgi vesicle-mediated transport by GTP hydrolysis' created to represent? There's a lot of steps at which GTP hydrolysis works -budding, docking, etc. This term seems both too vague and too specific (only for Sar1?). 3) Is it possible to make the current synonym 'cargo loading into COPII vesicle' the main term name for GO:90110 'COPII coat-cargo complex assembly' because it is more general? 4) Should GO:90110 be a part_of child of GO:90114 'COPII vesicle budding'? I suppose this depends on whether or not you consider cargo loading intrinsically tied to coat assembly. We are inclined to think so, but could be convinced otherwise.

Thanks! -Julie

Original comment by: juliep

[gocentral]

Hi Julie et al,

Tanya told me that I could go ahead and answer these so here goes:

1) Could the CC term 'COPII vesicle coat' GO:30127 get some sort of parentage up to ER?

We were actually more tempted to remove the golgi vesicle parent. I think that it would be better to simply have this an is_a 'vesicle coat' because of the lineage of its parent, 'ER to Golgi transport vesicle' . Also, this way, we can go back and create generic and specific terms for the creation of vesicles, the 'ER to Golgi transport vesicle' being one of them. See the answer to question #1 below. What do you think?

2) Please add the following synonyms for GO:90114 'COPII vesicle budding' -ER exit (narrow) DONE -COPII vesicle formation (exact) NOT DONE. Instead, I created a new term called 'COPII-coated vesicle formation' and this will be a part_of it. We would also need a term for the final pinching off of the vesicle. See the answer to #1 below. What do you think?

-ER vesicle budding (exact? broad?) DONE/BROAD (there are bound to be others we don't know about)

3) Please add the following synonyms for GO:48208 'COPII coating of ER-derived vesicle' -COPII vesicle coat formation (exact) DONE -ER vesicle coat formation (exact? broad?) DONE/BROAD

4) Merge GO:48221 'rough ER to cis-Golgi vesicle-mediated transport' into GO:6888 'ER to Golgi vesicle-mediated transport'. These two terms seem to be redundant. DONE (we also wanted to do this)

5) We would like a new term 'regulation of COPII vesicle coating' synonym: 'regulation of COPII coating of ER-derived vesicle' (exact) standard defs and parentage. We're not sure if it can be a direct child of GO:60628 or whether you need to create a new term 'regulation of COPII vesicle budding' to be a parent of this term, but we leave it up to you. DONE (did not create the budding term. Our general policy is to not create a grouping parent term unless there are 2 or more children to put under it. It just saves us work.)

Questions: 1) Other than Liz Miller, were there any other speakers who talked on secretion for whom we can look up their papers?

Not specifically on secretion, but there were lots of talks on budding and membrane dynamics that have to do with fusion and fission of membranes. There was some really cool stuff looking at single molecules. Unfortunately, I don't take down the names of people when I listen to talks. The session that touched on it the most was Minisymposium #6 on Sunday also see Symposium #2 on Sunday, the talk by Nunnari. He talked about the what was going on in the membranes as mitochondria fuse and divide. That's why we tended to want to go with the idea that there are two parts to making a vesicle the deformation of the membrane and then the actual membrane topological event that pinches off the bud.

2) For what exactly was GO:90113 'regulation of ER to Golgi vesicle-mediated transport by GTP hydrolysis' created to represent? There's a lot of steps at which GTP hydrolysis works -budding, docking, etc. This term seems both too vague and too specific (only for Sar1?).

Well, it seemed to be a general theme at the meeting that GTP hydrolysis was a key event in regulating all kinds of membrane dynamic events. Almost every talk about membranes touched on GTP hydrolysis and Liz Miller's talk focussed on it. We figure that this term, although broad as you point out, will allow us to describe not only the molecules that have the GTPase activity in which the GTP-GDP cycle seems to control the processes, but also the molecules that regulate the processes by regulating the GTP hydrolysis. The hope is that we will add even more children of this to describe exactly which steps in the process are regulated by GTP hydrolysis.

3) Is it possible to make the current synonym 'cargo loading into COPII vesicle' the main term name for GO:90110 'COPII coat-cargo complex assembly' because it is more general? DONE, but have chosen 'cargo loading into COPII-coated vesicle' since we hope to make the cellular component term.

4) Should GO:90110 be a part_of child of GO:90114 'COPII vesicle budding'? I suppose this depends on whether or not you consider cargo loading intrinsically tied to coat assembly. We are inclined to think so, but could be convinced otherwise.

LOL. We did this and then undid it then did it then undid it and in the end we couldn't decide whether it was a part_of the process. You have convinced us that it should be.

Original comment by: ukemi

[gocentral]

Actually, I didn't do:

COPII vesicle formation (exact) NOT DONE. Instead, I created a new term called 'COPII-coated vesicle formation' and this will be a part_of it. We would also need a term for the final pinching off of the vesicle. See the answer to #1 below. What do you think?

yet. I'd like to get your feedback first!

David

Original comment by: ukemi

[gocentral]

Hi David and Tanya,

I'll try to use the same numbering sections as below to try and keep things clear.

Edits: 1) Removing the Golgi parent for COPII sounds fine to us.

2) We're really uncomfortable with making separate terms for formation and pinching for the vesicles. The situation for organelles and that for vesicles may be different because at least for COPII, the whole budding/formation/pinching all happen together. We reviewed the entire body of literature for SEC13/31 in cerevisiae and they are intrinsically tied. But do you have examples of specifics in other organisms where they are separated?

3-5) Thanks!

Questions: 1) See answer to #2 above.

2) While GTP hydrolysis is a general theme for regulation of this process, we also think that this is really not that novel or such a big deal that this needs it's own term with "GTP hydrolysis" in the name. Why specifically separate this out? There are a lot of processes that are regulated by GTP hydrolysis, for example, translation is regulated by GTP hydrolysis. If you think the term should be kept then we think that all the regulation terms in this branch (at least so far) should be children of the GTP-hydrolysis regulation term. Does that make sense?

3) Cool.

4) LOL That's awesome.

Thanks again! -Julie and co.

Original comment by: juliep

[gocentral]

Hi Julie and co!

Wrt not having separate terms for the xxx vesicle formation vs. xxx vesicle budding, I looked at a couple of papers from Arabidopsis and Plasmodium and don't see evidence (yet) of these being separate so I'm ok with not splitting these out.

In response to the Question #2 discussion, which I'll just copy the last bit on:

>2) While GTP hydrolysis is a general theme for regulation of this process, >we also think that this is really not that novel or such a big deal that >this needs it's own term with "GTP hydrolysis" in the name. Why >specifically separate this out?

There are other ways to regulate ER to Golgi vesicle-mediated transport besides by GTP hydrolysis. For example, in this paper, the authors discuss how cargo loading regulates that same process.

http://jcs.biologists.org/cgi/content/abstract/121/18/3052

So, having the specific term including the 'by GTP hydrolysis' gives more specificity for annotation.

>There are a lot of processes that are >regulated by GTP hydrolysis, for example, translation is regulated by GTP >hydrolysis. If you think the term should be kept then we think that all >the regulation terms in this branch (at least so far) should be children of >the GTP-hydrolysis regulation term. Does that make sense?

Yes, there are many, many processes that are regulated by GTP hydrolysis. As David reminded me, other talks at the meeting emphasized how GTP hydrolysis is very much involved in the regulation of membrane dynamics (dynamin and similar proteins). We will probably end up adding more specific child terms to existing regulation terms (or improving existing ones) as the biology in these areas becomes more clear.

I hope that makes sense.

I'm also looking through my notes to see if I wrote down the names of people (besides Liz Miller) who talked related to secretion in yeast that I attended. I've only got this one that David mentioned already:

Jodi Nunnari, UC Davis - mitochondrial division and fusion (yeast)

So many concurrent talks!

Tanya

Original comment by: tberardini

[gocentral]

Hi Tanya and David,

So since it sounds like that you are going to keep the GTPase term, are you going to make the other terms in this branch with the words "regulation of" (specifically "regulation of COPII vesicle coating") children of this term?

>There are other ways to regulate ER to Golgi vesicle-mediated transport >besides by GTP hydrolysis. For example, in this paper, the authors discuss >how cargo loading regulates that same process. >http://jcs.biologists.org/cgi/content/abstract/121/18/3052 >So, having the specific term including the 'by GTP hydrolysis' gives more specificity for annotation.

We see your point that there are other regulatory nodes. But then by this argument, "cargo loading" should be some sort of child of regulation of these processes. However, there are gene products that are involved in cargo loading that are not considered regulators of the pathway. It's just for us as annotators who have read through this field, the GTP hydrolysis regulation term doesn't make sense to use.

Thanks for letting us know about Jodi Nunnari. It's actually quite convenient for us since Edith was a grad student in her lab. =)

regards, -Julie and posse

Original comment by: juliep

[gocentral]

Hi Julie and posse,

>So since it sounds like that you are going to keep the GTPase term, are >you going to make the other terms in this branch with the words "regulation >of" (specifically "regulation of COPII vesicle coating") children of this >term?

Yes. Based on the the paper that Tanya pointed out we would make a term like 'regulation of cargo loading' which would be an is_a child to the generic regulation term.

Sorry we are being so persnickity about the GTP hydrolysis stuff, but many of the speakers mentioned GTP hydrolysis in these types of processes. In the mitochondrion talk, it was in the context of dynamin, but that was not the main point of the talk. There was another talk where they talked about dynamin and also pointed out the importance of GTP hydrolysis. As far as using the terms for annotation, remember that we are talking about the process of GTP hydrolysis and not the function of GTPase activity. So molecules that regulate the hydrolysis of the GTP like GAPS would also be annotated to this, in addition to the molecule that has the GTPase activity. Would you be more comfortable if we made the GTP hydrolysis a part_of the processes and the regulation of the GTP hydrolysis regulate the processes? This distinction to make it a regulation rather than a part_of was our subjective decision based on the way we thought people were treating it with respect to the processes themselves. Tanya found a weed paper where I'm pretty sure she would have used this term, or at least had evidence that there was a gene product that would be annotated to it.

>Thanks for letting us know about Jodi Nunnari. It's actually quite >convenient for us since Edith was a grad student in her lab.

Cool, can she have a look at the mitochondial fission/fusion terms? I think we wanted to add a few more, but would love comments on what's there. While we are doing this, we might as well get it right.

D & T

Original comment by: ukemi

[gocentral]

Hello again,

I've been thinking about annotation. :) When we were adding these terms to the ontology we were certainly thinking that they'd be used for annotation and that we would use them for annotating genes from either mouse or Arabidopsis should we run across papers that described these processes in these organisms.

So the sticky term is

'regulation of ER to Golgi vesicle-mediated transport by GTP hydrolysis'

So, I was thinking about when I would realistically use such a term in annotation.

Here's the situation that I came up with:

I have a mutant that either doesn't hydrolyze GTP (Sar1 mutant or GTP-locked Sar1) or can't activate the GTPase (Sec23 mutant). GTP hydrolysis cannot occur in either case. I observe that in neither of these mutants does COPII vesicle formation/ER to Golgi vesicle-mediated transport occur as in wild type. (lots of papers have shown this, review from 2007 PMID: 17316621). (Had to stick to yeast papers for this one)

I could use just the parent term

'regulation of ER to Golgi vesicle-mediated transport'

with an IMP evidence code and that would be correct.

However, I could convey more information by using the term with 'by GTP hydrolysis' because I know that GTP hydrolysis does not occur in the situations described above (GTPase is dead or neutralized, or GTPase cannot be activated). I would still use the IMP evidence code.

Is that just too far out there? I think it makes sense but am always willing to discuss and hear other people's input.

More cents,

Tanya

Original comment by: tberardini

[gocentral]

Hey all,

Biologically speaking, I agree with the fact that GTP hydrolysis has a role in regulating these processes.

However, for me, having a term name containing 'by GTP hydrolysis' seems like a philosophical shift in the type of information being captured in the ontology. Not saying that we shouldn't do it but it seems like a greater discussion is needed first. Particularly about whether or not the mechanism of regulation or energy production (GTP vs. ATP) should be captured as part of the term and how to do this consistently for annotators who are suggesting new terms. What happens when two organisms use different mechanisms? Is this useful?

For SAR1, we are planning to annotate to the new term 'regulation of COPII vesicle coating' (GO:0003400). But if you want to keep the 'regulation of ER to Golgi vesicle-mediated transport by GTP hydrolysis' (GO:0090113) term, maybe this should be the parent for 'regulation of COPII vesicle coating'?

eurie

Original comment by: eurie

[gocentral]

Revisiting this term: 'regulation of ER to Golgi vesicle-mediated transport by GTP hydrolysis'

We will leave the term in for now and if it is never used for annotation in forseeable future, we will remove it as we create function process links. The anticipated function process link is a ways off but would probably look something like this:

GTPase activity involved in regulation of ER to Golgi vesicle-mediated transport

Original comment by: tberardini

[gocentral]

Hi Tanya,

I just came across this a bit randomly, and I wanted to comment about the example annotation you had for 'regulation of ER to Golgi vesicle-mediated transport by GTP hydrolysis':

in the situation you describe, the relationship between GTP hydrolysis and transport is not shown; you just have two independent pieces of data : (1) the mutant doesn't hydrolyze ATP, and (2) there is no ER to golgi transport. But you have no evidence that the ATP hydrolysis takes place at the actual step of transport. It could be that the GTPase function has nothing to do with the transport function of the protein; for example it could just be an earlier activation step depending on the substrate, (for example).

The problem with those terms is that annotations usually require multiple independent pieces of data - so it's some sort of IC?

... does that make sense?

Pascale

Original comment by: pgaudet

[gocentral]

Hi Julie,

Can we close this? I think we're done or have I missed something?

Thanks for checking too,

Tanya

Original comment by: tberardini

[gocentral]

Hi Tanya,

I think there's just one open question left (for us at least).

We thought that GO:0003400 : regulation of COPII vesicle coating should probably be a child of GO:0090113 regulation of ER to Golgi vesicle-mediated transport by GTP hydrolysis. Your thoughts on this?

Thanks! -Julie

Original comment by: juliep

[gocentral]

Hi Julie,

Yes, I agree and will make that update.

Thanks again,

Tanya

Original comment by: tberardini

[gocentral]

• status: open --> closed-fixed

Original comment by: tberardini