NTR: terms for alpha-glucosidase activity for different linkage substrates

[gocentral]

This paper (PMID: 24649402) shows different substrate specificities of 4 isomaltases (IMA1, IMA2, IMA3, IMA5). Specifically, the enzymes are only active on alpha-D-glucopyranosides. They show the following activities in various combinations:

alpha-1,6-glucosidase activity alpha-1,2-glucosidase activity (sucrose, kojibiose) alpha-1,3-glucosidase activity (nigerose, turanose) alpha-1,5-glucosidase activity (leucrose)

All of these IMA genes are currently annotated to oligo-1,6-glucosidase activity, but I would like to include the other activities when appropriate.

The "alpha glucosidase activity" GO term (GO:0004558) is specific to alpha-1,4 linkages (Definition: "Catalysis of the hydrolysis of terminal, non-reducing alpha-(1->4)-linked alpha-D-glucose residues with release of alpha-D-glucose"). Would all of these then go under its parent, "glucosidase activity," instead? There is a "glucan 1,3-alpha-glucosidase activity," but it is a child of "alpha-glucosidase activity" (unless I am not understanding, it seems incorrect to have a 1,3-alpha link under something with a 1,4-alpha definition?).

Please let me know the best way to sort this out.

Many thanks! Selina

The various substrates they use for demonstrating activity are shown above next

Reported by: ssdwight

Original Ticket: geneontology/ontology-requests/10984

[gocentral]

• assigned_to: Tanya Berardini

Original comment by: tberardini

[gocentral]

Hi Selina,

"There is a "glucan 1,3-alpha-glucosidase activity," but it is a child of "alpha-glucosidase activity" (unless I am not understanding, it seems incorrect to have a 1,3-alpha link under something with a 1,4-alpha definition?)"

You're right, this definitely needs fixing. It's also a bit tricky because as you've seen, the term is more broad than its definition. If we rename the term according to the def (most likely situation), annotations made directly to this term will need to be revisited to make sure that there's no misannotation to the more specific meaning.

We'll talk about it next editors' meeting.

In the meantime, I will remove the wrong parentage and move all the non-1,4 terms up to under the parent glucosidase for now. I'm leaning toward creating a new umbrella alpha-glucosidase term to cover all types of linkages (1,2 1,3 1,4 and 1,6)that when cleaved release an alpha-D-glucose.

As for the terms that you need for your annotations, I'd suggest creating new ones to describe the individual reactions with the specific substrates that were shown in the experiments. Can you give it a shot and write up proposed defs?

This may take a little while but we'll get there.

Thanks,

Tanya

Original comment by: tberardini

[gocentral]

Fixed incorrect parentage. Moved terms up to be under 'glucosidase activity' for now. More work to be done.

Original comment by: tberardini

[gocentral]

Editors' meeting consensus was to make the definition of "alpha glucosidase activity" GO term (GO:0004558) more generic and bring the child terms back. From your initial request, it sounds like you'll need:

alpha-1,6-glucosidase activity alpha-1,2-glucosidase activity, acting on sucrose alpha-1,2-glucosidase activity, acting on kojibiose alpha-1,3-glucosidase activity, acting on nigerose alpha-1,3-glucosidase activity, acting on turanose alpha-1,5-glucosidase activity, acting on leucrose

How about defs along the lines of:

Catalysis of the hydrolysis of terminal, non-reducing alpha-(1->6)-linked alpha-D-glucose residue with release of alpha-D-glucose.

Catalysis of the hydrolysis of the terminal, non-reducing alpha-(1->2)-linked alpha-D-glucose residue of sucrose with release of alpha-D-glucose.

Does the 'terminal, non-reducing alpha-(1->n)' part make sense for all of these?

Original comment by: tberardini

[gocentral]

Hi Selina,

Have you had a chance to review the suggestions here? I do need your feedback on this question. (see comment above for context) After that, I'm ready to proceed with the edits.

"Does the 'terminal, non-reducing alpha-(1->n)' part make sense for all of these?"

Thanks,

Tanya

Original comment by: tberardini

[gocentral]

Message from July 14, 2014

Hi Tanya,

Thanks so much for taking a look at this. Yes, I was hoping for a suggestion like this if it made sense with this area of the ontology - I can definitely create new term requests. Should I make a new SF ticket for each term or just keep them in this NTR string?

Thanks! Selina

Original comment by: ssdwight

[gocentral]

Message from July 14:

Hi Tanya,

It looks like “alpha-glucosidase” is the same as alpha-1,4-glucosidase:

http://en.wikipedia.org/wiki/Alpha-glucosidase

According to the EC numbers, alpha-glucosidase (3.2.1.20) and oligo-1,6-glucosidase (3.2.1.10) fall under the glycoside hydrolase (3.2.1.-) family (http://www.cazy.org/Glycoside-Hydrolases.html).

glucan 1,3-alpha-glucosidase is 3.2.1.84, so it looks like it should be directly under glycoside hydrolase (= glucosidase), as you have now moved it.

So, alpha-glucosidase seems like it represents a more specific action than its name implies - I think using the term glucosidase as the parent for the different linkages will work, but maybe we can’t use alpha-glucosidase as the umbrella term?

I am going to write the authors and see if they have a suggestion for the specific activity given the substrates they suggested - I am finding it difficult to decipher from the enzyme nomenclature.

Thanks! Selina

Original comment by: ssdwight

[gocentral]

Message from July 24:

Hi Tanya,

Thanks for bringing this up in the editor’s meeting. I’m sorry, we’ve been swamped with new page testing and getting ready for a meeting, so I am just revisiting this.

What you propose seems sensible to me, but after writing the author, I am now confused. Here are the terms I would use based the EC numbers he provided (his notes pasted at end of email):

alpha-1,6 EC 3.1.2.10 http://enzyme.expasy.org/EC/3.2.1.10 oligo-1,6-glucosidase (synonym is oligosaccharide alpha-1,6,glucosidase) Right now, oligo-1,6-glucosidase (GO:0004574) is a child of "glucosidase activity." All of our IMA genes are annotated to this GO term and are associated with the above EC number. Would adding them to a new term, alpha-1,6-glucosidase activity make sense?

sucrose (alpha-1,2, from my understanding of the paper) EC 3.2.1.26 (Invertase in CAZy database, glucose-β D fructose substrate) http://enzyme.expasy.org/EC/3.2.1.26 Beta-fructofuranosidase This is GO:0004564 (synonym invertase) (I’m a little confused as to why he matched this as it’s definition has to do with "hydrolysis of terminal non-reducing beta-D….” - I think I will write him back about this one)

alpha-1,3 He says alpha 1,3 goes with EC 3.2.1.27, but this is a deleted entry in ExPASy (here it is in BRENDA: http://www.brenda-enzymes.org/php/result_flat.php4?ecno=3.2.1.27), and we don’t have a similar term in GO unless we consider this one, which I believe is the one you moved up: GO:0033919 glucan 1,3-alpha-glucosidase activity maps to 3.2.1.84, but this is for glucans, not a disaccharide: http://enzyme.expasy.org/EC/3.2.1.84 Maybe an alpha-1,3-glucosidase terms would be good to have for this one. I do not need this term based on his feedback, but passing along the EC number he provided

alpha-1,4 He says that I don’t need the alpha-1,4 activity, but would put this with 3.2.1.20, which is alpha-glucosidase in ExPASy: http://enzyme.expasy.org/cgi-bin/enzyme/enzyme-search-ec This is the same as GO:0004558 = alpha-glucosidase activity and GO and EC basically have the same definition.

So, I now know where to annotate based on the ECs he provided, but the question is whether the GO hierarchy makes sense and whether we want the GO terms to match the EC names? For instance, if we move alpha-glucosidase to be an umbrella term, it will no longer match the EC's alpha-glucosidase, which is where the 1,4 linkage definition was derived. Also, there seems to be a mix of oligo- and glucan- vs. just alpha -, so maybe it would help to clear that up.

Let me know what you think based on all of this. I am not that familiar with GO/EC mappings. For now, I can certainly use the GO terms based on mapping to the EC terms he provided.

Thanks! Selina

Original comment by: ssdwight

[gocentral]

Message sent on August 14:

Hi Tanya,

I sent a mail back on July 24 after corresponding with the author - will forward to you.

Thanks! Selina

Original comment by: ssdwight

[gocentral]

Hi Selina,

Thanks for adding in all the comments that went missing.

"alpha-1,6 EC 3.1.2.10 http://enzyme.expasy.org/EC/3.2.1.10 oligo-1,6-glucosidase (synonym is oligosaccharide alpha-1,6,glucosidase) Right now, oligo-1,6-glucosidase (GO:0004574) is a child of "glucosidase activity." All of our IMA genes are annotated to this GO term and are associated with the above EC number. Would adding them to a new term, alpha-1,6-glucosidase activity make sense?"

I don't think we need a new term, GO:0004574 should suffice. From the Metacyc page for this reaction, I think this is the right way to go.
http://biocyc.org/META/NEW-IMAGE?type=REACTION&object=3.2.1.10-RXN&redirect=T I added 'releases a free alpha-D-glucose' to the definition.

Original comment by: tberardini

[gocentral]

"sucrose (alpha-1,2, from my understanding of the paper) EC 3.2.1.26 (Invertase in CAZy database, glucose-β D fructose substrate) http://enzyme.expasy.org/EC/3.2.1.26 Beta-fructofuranosidase This is GO:0004564 (synonym invertase) (I’m a little confused as to why he matched this as it’s definition has to do with "hydrolysis of terminal non-reducing beta-D….” - I think I will write him back about this one)"

Did you hear anything back on this one?

Original comment by: tberardini

[gocentral]

OK. I think I've got it now. I'm attaching a shot of what it looks like now. I hope this covers all the terms you need. I'll wait to hear back on the sucrose alpha1,2 one.

Original comment by: tberardini

[gocentral]

Hi Tanya,

Thanks for working on this. It seems to make sense to me with the new definition for alpha-glucosidase activity. I only have a couple of questions:

1) For glucan 1,4-alpha-glucosidase activity, there is a Comment: Note that this term is not a child of 'alpha-glucosidase activity ; GO:0004558', because in the reaction represented by GO:0004339 results in the release of beta-D-glucose, whereas in GO:0004558 alpha-D-glucose is released.

Is this OK to have as a grandchild of alpha-glucosidase activity?

2) glucan 1,6-alpha-glucosidase activity and oligo-1,6-glucosidase activity are both children of alpha-glucosidase activity (e.g. siblings) This seems somewhat different than the other linkage relationships at first glance.

I'm getting a little confused because I want to make the GO nomenclature intuitive, but I am not sure that matches the enzyme descriptions in other databases?

Selina

Original comment by: ssdwight

[gocentral]

Hi Selina,

1 - Note that this term is not a child of 'alpha-glucosidase activity ; GO:0004558', because in the reaction represented by GO:0004339 results in the release of beta-D-glucose, whereas in GO:0004558 alpha-D-glucose is released.

You are totally correct. I've moved the term to be a direct child of glucosidase activity and modified the comment to read as follows, with the correct GO id in place.

Note that this term is not a child of 'alpha-glucosidase activity ; GO:0090599', because in the reaction represented by GO:0004339 results in the release of beta-D-glucose, whereas in GO:0090599 alpha-D-glucose is released.

1. It seems to me that oligo-1,6-glucosidase is ok here as the activity releases alpha-D-glucose. I'm not sure from looking at the papers whether glucan-1,6-alpha-glucosidase releases alpha-D-glucose or not, all I see right now is the release of D-glucose. To be on the safe side, I'll move the glucan term out to be directly under glucosidase activity.

We do our best with the naming and placement. As you know, there's always room for improvement and we're ready to make changes anytime to accommodate users' needs.

Thanks for your help,

Tanya

P.S. I think we're almost done with this one. Just the issue of the 'alpha-1,2' left.

Original comment by: tberardini

[gocentral]

Hi Tanya,

Thanks for your quick reply!

It sounds like glucan-1,4- and glucan-1,6 will go directly under glucosidase - should this be the case for glucan-1,3 also? I can snoop around in the literature if needed.

I'll send another note to the author now.

thx! Selina

Original comment by: ssdwight

[gocentral]

glucan-1,3-alpha-glucosidase: Looks like this activity releases alpha-D-glucose, so I'd like to leave it under alpha-glucosidase.

glucan-1,4-alpha-glucosidase: Releases beta-D-glucosidase, so this can't be directly under alpha-glucosidase. The naming is confusing but refers to the (1,4)-linked α-D-glucose residues that are targeted.

glucan 1,6-alpha-glucosidase activity: As I said earlier, I can't tell which stereoisomer is released: alpha- or beta-D-glucose. Therefore, I'm going to leave this directly under glucosidase activity.

It's so nice and clean, the way the terms are named but not so clean once you get into the actual reactions and products. Sorry about that but I think splitting the terms into the above, separate locations is the way we can best capture our current knowledge about the activities.

Original comment by: tberardini

[gocentral]

Sounds good, Tanya. No worries - the hierarchy has to be driven by the reactions and definitions, not the term names, and it sounds like you have done this.

The attachment looks good. I'll go back and curate accordingly.

Thanks for working through this with me.

Selina

Original comment by: ssdwight

[gocentral]

Thanks to you too. I'll wait for the response from the author about the alpha-1,2.

Tanya

Original comment by: tberardini

[gocentral]

Hi Tanya,

Sure!

I am looking back on the alpha-1,2 - he gave the alpha-1,2 linkage for sucrose the EC# 3.2.1.26. This is equivalent to beta-fructofuranosidase activity = GO:0004564, which has the definition:

Hydrolysis of terminal non-reducing beta-D-fructofuranoside residues in beta-D-fructofuranosides

I've attached the substrate sucrose (reads O-α-d-Glucosyl-(1→2)-β-d-fructose in the paper). I think what is happening from looking at the structure is that there is an alpha-1,2 linkage, but the terminal residue is a beta-D-fructofuranoside? So, in this case, the beta-fructofurranosidase activity makes sense?

From this perspective, I would not need a new GO term for the alpha-1,2...

I actually think I can go to this term: GO:0004575 sucrose alpha-glucosidase activity, which is a child of beta-fructofuranoside and alpha-glucosidase activity.

What do you think?

thx! Selina

Original comment by: ssdwight

[gocentral]

Hi Selina,

You've hit on the right term! 'sucrose alpha-glucosidase activity' should do it for you.

If you think we've completed all the work you need on this issue, please close it. If not, just let me know what else you need.

Thanks,

Tanya

Original comment by: tberardini

[gocentral]

Great - thanks, Tanya, for all your help! Closing now.

Original comment by: ssdwight

[gocentral]

• status: open --> closed-works-for-me

Original comment by: tberardini