Improve consistency of labels of ABC transporter activities

[pgaudet]

Hello,

ATP-coupled transporters are named in several different ways:

• ATPase activity, coupled to movement of substances
• phospholipid-translocating ATPase activity
• ATPase-coupled anion transmembrane transporter activity
• vitamin B12 ABC transporter activity (and maybe other ways).

I will rename all these terms "ATPase-coupled transmembrane transporter activity" and "ATP-coupled transmembrane [x] transporter activity" for all children.

Thanks, Pascale

[pgaudet]

@amandamackie Does thsi term label look ok to you ?

[amandamackie]

Hi Pascale, I'm out of the office for next two weeks. Will look into this when I return. Amanda

Get Outlook for Androidhttps://aka.ms/ghei36

---

From: pgaudet notifications@github.com Sent: Wednesday, April 17, 2019 1:25:28 AM To: geneontology/go-ontology Cc: Amanda Mackie; Mention Subject: Re: [geneontology/go-ontology] Improve consistency of labels of ABC transporter activities (#17164)

@amandamackiehttps://protect-au.mimecast.com/s/1YexCjZ12RfyElZZsRYWeM?domain=github.com Does thsi term label look ok to you ?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://protect-au.mimecast.com/s/g2EPCgZ05JfY6P44uNOAWL?domain=github.com, or mute the threadhttps://protect-au.mimecast.com/s/jBS0Ck815RCqlYoocQO5E5?domain=github.com.

[vanaukenk]

@pgaudet - I have a question about GO:0004012, phospholipid-translocating ATPase activity.

Is that activity transmembrane transport, or does it refer to translocation of phospholipids from one monolayer to the other, e.g. cytoplasmic leaflet to exoplasmic leaflet?

[pgaudet]

Hi @vanaukenk Good point. These are described as flippases or scramblases, see https://en.wikipedia.org/wiki/ATP10A "The aminophospholipid translocases transport phosphatidylserine and phosphatidylethanolamine from one side of a bilayer to another."

I will not touch these for now - these will probably need an annotation review.

Thanks, Pascale

[ValWood]

should be placed under GO:0120013 intermembrane lipid transfer activity or GO:0140303 intramembrane lipid transporter activity which should also have a common parent, probably.

[ValWood]

corrected above to GO:0140303 intramembrane lipid transporter activity

but phospholipid-translocating ATPase activity (GO:0004012) appear to be defined as GO:0140303 intramembrane lipid transporter activity (inner to outer leaflet)

possibly equivalent?

the intermembrane ones are the sec14 cytosolic factor family

[amandamackie]

I have a couple of comments:

Transport ATPases are generally classified into 3 or 4 types: ABC (for ATP-binding cassette) transporters; P-type (they form a phosphorylated intermediate as part of the transport mechanism; see GO:0015662) and the F-type and V-type (the proton pumps; ?GO:0044769). Does this distinction need to be covered/preserved in molecular function terms? A good starting reference is PMID16691464

I noticed that the current term GO:0042626 (ATPase activity coupled to transmembrane movement of substances) is a child of GO:0015405 (P-P bond hydrolysis driven transmembrane transporter activity). This latter definition is misleading; it is the hydrolysis of phosphoanhydride bonds in ATP that provides energy for transport, not P-P bonds.

[pgaudet]

Thanks @amandamackie I dont think the parent term 'P-P bond hydrolysis driven transmembrane transporter activity' is needed anyway; I think it was intended for proteins that use another nucleotide as a source of energy, but I don't think this exists; does it ?

[pgaudet]

Changed label from 'ATPase activity, coupled to transmembrane movement of substances' to 'ATPase-coupled transmembrane transporter activity' added synonyms: +synonym: "ATP-coupled transmembrane transporter activity" EXACT [] +synonym: "ATP-dependent transmembrane transporter activity" EXACT [] +synonym: "ATPase activity, coupled to transmembrane movement of substances" EXACT []

Changed subclass from 'is a' ATPase activity, coupled to movement of substances to has_part GO:0016887 ! ATPase activity

@ukemi @cmungall Can I merge this (ie using 'has part' for compound functions') ? Pascale

[deustp01]

Transport ATPases are generally classified into 3 or 4 types: ABC (for ATP-binding cassette) transporters; P-type (they form a phosphorylated intermediate as part of the transport mechanism; see GO:0015662) and the F-type and V-type (the proton pumps; ?GO:0044769). Does this distinction need to be covered/preserved in molecular function terms?

Three or four different molecular mechanisms should be three or four different molecular functions, right? Especially as each is enabled by a large and distinct family of proteins.

[pgaudet]

Ontology call: I will go ahead with this branch (using 'has part' for ATPase activity of the ABC transporter, and 'transmembrane transporter activity' as the genera. )

[amandamackie]

Thanks @amandamackie I dont think the parent term 'P-P bond hydrolysis driven transmembrane transporter activity' is needed anyway; I think it was intended for proteins that use another nucleotide as a source of energy, but I don't think this exists; does it ?

I think GTP can be used as a source of energy for transport but my point was more that there are no P-P (ie phosphorous-phosphorous) bonds in ATP or any other nucleotide, the high energy bonds are phosphanhydride bonds which are the bonds between two phosphate molecules. I guess it's a non-issue if the term is being removed.

[pgaudet]

https://github.com/geneontology/go-ontology/pull/17220

[pgaudet]

I also need to check EC cross references. @hdrabkin I see that some of the ECs you've added from pz; was this some time ago ? Many of these have new ECs (7.x.x.x). Were you planning to change these? (I can do it; I just don't want to duplicate work).

Thanks, Pascale

[hdrabkin]

Yes, that is why I put the EC:7.x terms in for parents; But, I have NOT changed ones that I could not assign correct RHEA to (usually due to specificity issues, etc. However, I can go in and just change the EC numbers. Let me go through my spreadsheet for these

From: pgaudet notifications@github.com Reply-To: geneontology/go-ontology reply@reply.github.com Date: Monday, May 6, 2019 at 4:10 PM To: geneontology/go-ontology go-ontology@noreply.github.com Cc: me Harold.Drabkin@jax.org, Mention mention@noreply.github.com Subject: Re: [geneontology/go-ontology] Improve consistency of labels of ABC transporter activities (#17164)

I also need to check EC cross references. @hdrabkinhttps://github.com/hdrabkin I see that some of the ECs you've added from pz; was this some time ago ? Many of these have new ECs (7.x.x.x). Were you planning to change these? (I can do it; I just don't want to duplicate work).

Thanks, Pascale

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/geneontology/go-ontology/issues/17164#issuecomment-489758367, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ACPU3EDM6R7KJNYTJ464OOTPUCGDHANCNFSM4HGKFTFQ.

The information in this email, including attachments, may be confidential and is intended solely for the addressee(s). If you believe you received this email by mistake, please notify the sender by return email as soon as possible.

[hdrabkin]

I see 7 on my list; all of these had rhea issues so I did not edit the term but I can go in and just change the EC right now. This is why I added the EC:7.n parent classes because the is_a will need to change to map to the new EC class. example

GO:0102017 | ATPase-coupled alkylphosphonate transmembrane transporter activity | EC:3.6.3.28 transferred entry: 7.3.2.2 |

[hdrabkin]

Ah; found another "L-arginine-importing ATPase activity EC:3.6.3.21 Transferred entry: 7.4.2.1 Will do this under my issue for the rhea stuff

[pgaudet]

Great ! I have been using the redirects from EC; I hope these are OK.

[pgaudet]

@hdrabkin WRT 'L-arginine-importing ATPase activity': I was going to propose to obsolete these (or merge into the parent), since the 'importing' part is a process, not a function.

Can I add this as another xref to the parent term even if it's not the exact MF ? In other words, are we trying to be exhaustive about xrefs to EC ?

Thanks, Pascale

[deustp01]

WRT 'L-arginine-importing ATPase activity': I was going to propose to obsolete these (or merge into the parent), since the 'importing' part is a process, not a function.

Hmm. Is the word "importing" used here to indicate the overall process of arginine uptake, or is it used narrowly to indicate the direction of the arginine transport activity (into the cell across the plasma membrane)?

[hdrabkin]

Yes; but make sure if you do that you change the is_a parent; Tomorrow I'll add the 3 digit EC for class 7 (I already added the 1 and 2 digit. I

[hdrabkin]

7.4.2.1 ABC-type polar-amino-acid transporter I think it should be fine as parent is arginine-importing ATPase activity, and arginine is polar.

[pgaudet]

Is the word "importing" used here to indicate the overall process of arginine uptake, or is it used narrowly to indicate the direction of the arginine transport activity (into the cell across the plasma membrane)?

What would be the difference ? Isn't that a process regardless?

[hdrabkin]

I would think import does imply a direction. The EC def "ATP + H2O + polar amino acid-[polar amino acid-binding protein][side 1] = ADP + phosphate + polar amino acid[side 2] + [polar amino acid-binding protein][side 1] " doesn't specifiy what side is outside or in. However, the RHEA is very explicit "ATP + H2O + L-arginine(out) = ADP + H(+) + L-arginine(in) + phosphate Cross reference: EC 7.4.2.1 So I would not remove the import from the term name

[deustp01]

It's not a process because the transporter in a single step mediates the translocation of a molecule of arginine inwards across the plasma membrane simultaneously with the hydrolysis of a molecule of ATP, so no more a process than any other ATP-dependent transport activity.

[pgaudet]

If it's not a process then most 'import' and 'export' terms must be removed from Process fo eg:

For eg children of GO:0098739 import across plasma membrane

Most are single steps....

[deustp01]

If by definition processes cannot be single steps (maybe not a universally popular definition), then as you say most of the terms on the list are functions, not processes.

[hdrabkin]

Just looking; if you merge the L-arginine importing into L-arginine transmembrane transporter, the EC might map (side1/side2), but the RHEA would no longer map correctly because the RHEA specifies direction.

[ValWood]

But many processes are the sum of many different transporters.

"Glucose import" for example can be performed by many different gene products of different affinities high, low, of expressed under certain conditions. The process of glucose import does therefore have multiple steps they just happen to be performed (largely) by gene products doing the same MF.

Import and export are also regulated by different pathways, so it is important to distinguis at the process level.

So importer is a function, but import is a process.

My preference would be to have import as a process, because helps us when describing processes to know the directionality of transport.

I guess its an arbitrary decision and we could have both MF and BP, but I would really like to keep import and export biological processes for modularization. For example import and export into the vacuole /mitochondria/ across plasma membrane are differently aligned with other processes. For many things if we only capture as "transport" we lose a lot at the process level.

Also, we can easily couple the basic "transporter" MF to the import process to define the specificity here is a recent nice example paper using zinc: https://www.pombase.org/reference/PMID:29529046

[ValWood]

Please don't remove import/export from process...;(

[ValWood]

If we did do this this, to get the specificity in the zinc example above we would need to do

for zinc ion transmembrane transporter activity- >

would need zinc ion organellar transmembrane importer activity zinc ion vacuolar transmembrane importer activity etc

it would explode the MF branch but would not gain anything.....

and everyone who has done merges would need to reannotate again

[ValWood]

@ukemi phew!

[ukemi]

PMID:23908456

[pgaudet]

@ValWood Right, I was hoping to keep transmembrane transporter activity in MF and import/export in BP (this is what you and I discussed, as far as I remember).

But I wanted to know how @deustp01 was envisioning the split.

Pascale

[deustp01]

PMID:23908456 from @ukemi illustrates the split really nicely for glutamine: glutamine transport is an activity mediated by a specific gene product or gene products in the plasma membrane of specific kinds of kidney cells, but it occurs as one part of a larger process (details of the other parts and their interconnections given in the paper), in this case tightly integrating several transport and metabolic activities to maintain pH.

I guess that a simpler but similar story could be told for arginine: there's a transporter activity (in this case, an ATP-dependent one), but in normal cells its function will always be part of a larger process in which other activities move counter-ions and other stuff to maintain charge balance and pH, whether or not we have the experimental data right now to make that complete process annotation.

In other words, arginine transport process looks like it is single-step only because we don't have the information at hand to fill in a full description of its other parts.

Does that help?

[pgaudet]

I think so - but import versus export should be kept under BP; right ? In any case - L-arginine-importing ATPase activity is now under https://enzyme.expasy.org/EC/7.4.2.1

The Accepted Name: is ABC-type polar-amino-acid transporter. Alternative Name(s): Histidine permease. Polar-amino-acid-transporting ATPase.

So a merge does seem appropriate.

Pascale

[hdrabkin]

The RHEA that goes with that EC will now not be correct if you merge the terms because the RHEA id is specific for arginine. https://www.rhea-db.org/searchresults?q=EC%3A7.4.2.1 There is NO generic RHEA . What can happen Term 1; xref of 7.4.2.1 (This entry comprises bacterial enzymes that import His, Arg, Lys, Glu, Gln, Asp, ornithine, octopine and nopaline. ) then we might end up with Term 1 a Same EC but has RHEA xref for arginine Term 1 b Same EC but has RHEA xref for lysine .... etc. Note that when I added RHEA to "L-arginine-importing ATPase activity I did NOT add that RHEA to its parent.

[hdrabkin]

I suppose I could request a RHEA entry that uses "polar amino acid zwitterion" , ChEBI ID | CHEBI:62031, as one of the substrates, if it's allowed. I assume the zwitterion is the pH 7.3 entity (although at pH 7.3, I would expect lysine R group to be further protonated for an additional positive charge, Glu would have an additional negative charge, etc. )

[hdrabkin]

Ah; I see RHEA:14674 a polar amino acid(out) + ATP + H2O => a polar amino acid(in) + ADP + H(+) + phosphate So may be ok.

[pgaudet]

Merge the following 2 terms (no annotations, no references): 'GO:0033224 2-aminoethylphosphonate transmembrane transporter activity' -> no xref merge into 'GO:0033225 ATPase-coupled 2-aminoethylphosphonate transporter activity' -> RHEA:32775 This is the only reaction in RHEA that involves 2-aminoethylphosphonate, suggesting this is the only type of transporter.

[pgaudet]

Obsoleted xenobiotic transporters (no annotations; and this is not a function, since the xenobiotic transporters are no specific):

id: GO:0103113 -name: glucosyl-oleandomycin-exporting ATPase activity +name: obsolete glucosyl-oleandomycin-exporting ATPase activity

id: GO:0043216 -name: ATPase-coupled daunorubicin transmembrane transporter activity +name: obsolete ATPase-coupled daunorubicin transmembrane transporter activity

[pgaudet]

Merge GO:0015623 ATPase-coupled iron-chelate transporter activity -> 0 annotations into its child GO:0015625 ATPase-coupled ferric-hydroxamate transmembrane transporter activity -> 2 EXP

both share the same EC, EC 7.2.2.16, that corresponds to GO:0015625.

[pgaudet]

Obsolete GO:0042969 lactone transport GO:0042971 lactone transmembrane transporter activity no annotations, no references, no xrefs.

[pgaudet]

Merge GO:0015604 organic phosphonate transmembrane transporter activity -> 3 IEAs, to ABC transporters (IPR005769 + IPR005770) and GO:0042917 alkylphosphonate transmembrane transporter activity -> 0 annotations, corresponds to the same genes annotated to GO:0015604 according to http://vm-trypanocyc.toulouse.inra.fr/ECOLI/NEW-IMAGE?type=POLYPEPTIDE&object=PHND-MONOMER&detail-level=2 into GO:0015416 ATPase-coupled organic phosphonate transmembrane transporter activity

@sarach06 @EBI-Hsinyu Does InterPro automatically map merged terms ?

Thanks, Pascale

[pgaudet]

And, merge corresponding processes:

GO:0042916 alkylphosphonate transport -> 0 annotation into parent GO:0015716 organic phosphonate transport

[pgaudet]

Merge GO:0070812 ATPase-coupled glycerol-2-phosphate transmembrane transporter activity -> 0 annotations, no references into GO:0015430 ATPase-coupled glycerol-3-phosphate transmembrane transporter activity

This is the same transporter, that can also transport G2P, see PMID:19429609

[pgaudet]

Merge

GO:0015595 ATPase-coupled spermidine transmembrane transporter activity -> 4 EXP into GO:0015417 ATPase-coupled polyamine transmembrane transporter activity -> 0 annotations

Same EC, same Rhea - the transporters are able to transport both substrates.

[pgaudet]

• GO:0015086 cadmium ion transmembrane transporter activity => 21 EXP, ABC-type, SLC
• GO:0015434 ATPase-coupled cadmium transmembrane transporter activity => 1 EXP annotation, TAIR HMA1 (AT4G37270), a P-type

• GO:0008551 cadmium transmembrane transporter activity, phosphorylative mechanism => 3 EXP annotations to P-types

  • [ ]  Note: RHEA:12132 corresponds to both ABC-type and P-type (currently linked to P-type). Emailed RHEA helpdesk.

[pgaudet]

Merge GO:0048474 D-methionine transmembrane transporter activity into GO:0033232 ATPase-coupled D-methionine transporter activity

I cannot find an example of a D-methionine transporter that is not an ABC transporter. (Also, it seems like they are all bacterial)

[pgaudet]

Obsolete GO:0033220 ATPase-coupled amide-transporter activity

-> unnecessary grouping class.