query about "cross-linking" and 'bundling'

[gocentral]

I notice that actin filament binding has the synonym 'actin cross-linking activity'

There also used to be a functionterm 'actin cross-linking activity' which is obsolete,

This term was made obsolete because it represents a combination of molecular functions. To update annotations, use the molecular function terms 'actin filament binding ; GO:0051015' and 'protein binding, bridging ; GO:0030674'.

but perhaps actin-cross linking is a valid function term which we would want to distinguish from actin binding

(perhaps 'actin cross-linking' could be a child of both actin filament binding and 'protein binding, bridging?)

Reported by: ValWood

Original Ticket: "geneontology/ontology-requests/2834":https://sourceforge.net/p/geneontology/ontology-requests/2834

[gocentral]

Logged In: YES user_id=516865

The paper I am reading PMID:11694585 appears to refer to cross-linking interchagably as 'actin cross-linking' and 'F-actin bundling activity' (p3518, section F-actin binding of fim1)

we have a term actin filament bundle formation, and this has the synonym actin bundling activity. This needs to also be considered if

The cross-linking term is reinstated. If it isn't it seems like further synonyms may be appropriate for some of these terms.

Original comment by: ValWood

[gocentral]

• summary: query about "cross-linking" --> query about "cross-linking" and 'bundling'

Original comment by: ValWood

[gocentral]

Logged In: YES user_id=631592

I think it is worthwhile to have a term called actin bundling or actin crosslinking. This term differs from simple actin binding in that it is a specific type of binding that bundles actin filaments together. There are also lots of other types of actin binding activities like capping and binding to actin for motile purposes.

The current term actin filament bundle formation is a process where actin filaments bcome parallel and form bundles. The molecular function of actin cross-linking activity would represent a molecule that acts as the linker between actin filaments.

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=473796

David, is actin bundling equivalent to actin crosslinking, or are they different?

Original comment by: girlwithglasses

[gocentral]

Logged In: YES user_id=631592

I think they can be used interchangeably, but actin crosslinking more precisely defines this activity.

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=516865

We have bundling in process, which seems to be appropriate because: Different proteins have cross-linking activity (fimbrin, villin, alpha-actinin, plastin, spectrin, filamin). Are these differnt specificities of cross-linking? and can a bundle therefore consist of different 'types' of cross-links (which would make it a multi-step process?)

David?

Original comment by: ValWood

[gocentral]

Logged In: YES user_id=631592

I think of the crosslinking activity as a molecular event that literally cross-liks two actin filaments. I don't think this can necessarily be broken down further without getting to a gene product level of things. I think of the process of bundle formation being not only this actual cross-linking, but any other activities that go into the formation of a bundle. For example when the bundle that creates the cleavage furrow forms, there is a lot more hapenning than just crosslinking of the filaments. We should get Rex's take on this as well.

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=516865

actin bundling activity OBSOLETE (was not defined before being made obsolete) This term was made obsolete because it represents a biological process. To update annotations, use the biological process term 'actin filament bundle formation ; GO:0051017'.

It sounds like we need to:

1. reinstate 'actin bundling activity' (as a child of i) actin binding and ii) protein binding, bridging becasue it is a specific type of both of these activities.

2. remove the synonym 'actin bundling activity' from 'actin filament bundle formation' (and perhaps extend the definition(s) to clarify the difference?)

Original comment by: ValWood

[gocentral]

Logged In: YES user_id=473796

I would like to put this term into process because it's two (or more) events of actin binding - i.e. the protein acting as the crosslink has to be linked to two or more filaments.

How about something like

actin crosslinking exact syn: formation of actin crosslink exact syn: actin bundling def: The process by which two or more actin filaments are bundled together by proteins that act as crosslinks between the filaments.

A better def would be much appreciated!

Original comment by: girlwithglasses

[gocentral]

Logged In: YES user_id=631592

Hi Amelia,

I can see your argument for there being two events of actin binding, but it still comes down to these events being strictly coupled and more than just actin-binding. For that reason, I still think that the actual cross-linking is a molecular function. Kinase activity is a function and the enzyme binds both ATP and the substrate. It comes back to the old argument of when does a function stop and a process begin, and I think we need to base that on the coupling of the activities. I heard back from Rex and he also thinks that actin crosslinking and actin bundling are not exactly the same thing. I am going to poke around in the literature some more to see what I can find.

David

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=473796

> I can see your argument for there being two events of actin > binding, but it still comes down to these events being > strictly coupled and more than just actin-binding.

What about if there was only one actin filament for the crosslinker to bind to? Presumably the crosslinker would still bind?

Further info on actin bundling and crosslinking would be great - thanks David!

Original comment by: girlwithglasses

[gocentral]

Logged In: YES user_id=631592

Yes, but that is not its normal function.

David

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=436423

I shall stay out of the function/process debate for the moment.

I have only one thing to contribute: actin bundling is most assuredly not exactly synonymous with actin crosslinking, because actin filaments can be crosslinked to form networks (which results in gel formation) as well as bundles; i.e., crosslinking is broader. In a bundle the actin filaments are all parallel or antiparallel.

m

Original comment by: mah11

[gocentral]

Logged In: YES user_id=631592

Excellent point Midori! According to Val's paper, some people use the terms interchangeably, so I think maybe it should be a non-exact synonym.

I have poked around a bit and it seems that most people believe that actin crosslinking is a function that occurs in the processes of bundling and network formation. Crosslinking proteins almost alway contain 2 F-actin binding sites and the bivalency is responsible for their crosslinking function. It appears that fascin and the Dicty 34kD protein are proteins with a cross-linking function that is involved in both bundling and network formation. The protein appears to function identically in both processes.

David

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=473796

So the current state of affairs is:

bundle formation - parallel / antiparallel actin filaments crosslinked together but all on the same axis

actin network formation - actin filaments crosslinked together, filaments not on the same axis, results in the formation of a "gel"

actin crosslinks - proteins that connect two actin filaments together, where the filaments may be on the same or differing axes

actin bundling - used by some synonymously with 'crosslinking', by others synonymously with 'bundle formation' --> GO's "related synonym" type

Is that right?

Original comment by: girlwithglasses

[gocentral]

Logged In: YES user_id=631592

Amelia,

This sounds good to me.

David

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=473796

have added 'actin filament network formation ; GO:0051639'

I can add 'actin crosslinking' but I think that it should be in process as it is two instances of protein binding. As you can have the crosslinker bound to just one actin filament (even if this isn't how it "should" be), I don't believe that the two binding events are interdependent and inseparable - yes, you need two filaments and a crosslinker for crosslinking to occur by definition, but I don't think that necessarily means that these conditions are equivalent to those in other function terms like those ATP-dependent transporter reactions.

Original comment by: girlwithglasses

[gocentral]

Logged In: YES user_id=473796

Does anyone have any objections to my adding actin crosslink formation/actin crosslinking to the process ontology? I'll put it in and we can close this item if so...

Original comment by: girlwithglasses

[gocentral]

Logged In: YES user_id=516865

so won't we have 'actin crosslinking activity' as a child of 'protein binding, bridging' ?

sorry!

Original comment by: ValWood

[gocentral]

Logged In: YES user_id=473796

Nope, we wouldn't sadly... but you can do a dual annotation, and I think in the future we will be able to make links between process and function, so these terms will be connected. Can't say fairer than that, can ya?

Original comment by: girlwithglasses

[gocentral]

Logged In: YES user_id=473796

Term added: actin crosslink formation ; GO:0051764

Original comment by: girlwithglasses

[gocentral]

Logged In: YES user_id=473796

Term added: actin crosslink formation ; GO:0051764

Original comment by: girlwithglasses

[gocentral]

• labels: --> New term request
• milestone: --> TermGenie
• assigned_to: nobody --> girlwithglasses
• status: open --> closed-fixed

Original comment by: girlwithglasses