search

geneontology / go-ontology

Source ontology files for the Gene Ontology
http://geneontology.org/page/download-ontology
Creative Commons Attribution 4.0 International
215 stars 39 forks source link

Taxon constraint: symbiont-containing vacuole membrane GO:0020005 #22162

Open Antonialock opened 2 years ago

Antonialock commented 2 years ago

Please provide as much information as you can:

Ancestor GO ID Ancestor GO Term Name Relationship Taxon ID Taxon Reference(s)
GO:0033643 host cell part Never in Taxon 7742 Vertebrata

"Irga6, Irgb6, Irgd, Irgm2 (GTPI) and Irgm3 have been shown to accumulate at T. gondii parasitophorous vacuoles (PVs). Accumulation of IRG proteins at the PVM is followed by vesiculation and disruption of the PVM and death of the parasite."

pgaudet commented 2 years ago

@dsiegele @mgiglio99 @genegodbold @cmungall any thoughts on that one?

This term is currently is_a host cell part Perhaps the solution is to move it directly under 'organelle'?

genegodbold commented 2 years ago

Well, this pathogen/parasite-containing vacuole (PCV) is the least host-cell-like of all the parts of the host cell. And this bit of cellular real estate has been hijacked and repurposed (usually by determining what Rab smallGTPases are and are not resident on the outer membrane of the vacuole) and is now actively harboring the enemy of the cell. Q: Is the host cell trying to re-establish control and exterminate the invaders by merging it with lysosomes? A: Certainly, but most intracellular bacteria (some still appear to be learning) have figured out ways to manipulate the host endosomal system so that this doesn't happen OR they sabotage the vacuolar ATPases so that they don't pump protons as efficiently so the acidic hydrolases fail to work.

Legionella pneumophila has 200-300 proteins that are devoted to controlling and regulating these parts of its host. That is probably more than mammalian cells use to regulate their endocytic traffic. That would actually be an interesting research question.

mgiglio99 commented 2 years ago

looping in @nsuvarnaiari

If you move it (and its parent) under organelle, that would imply an organelle of the organism being annotated which is fine for the host but not for the symbiont. Could it (and its parent) be moved to be a child of 'cellular anatomical entity' and take away all of the other parents? And remove the taxon constraint?

Antonialock commented 2 years ago

It is a bit of a weird case isn't it? Anyone against the proposed solution by @mgiglio99 to move it to be a child of 'cellular anatomical entity', take away all of the other parents, and remove the taxon constraint?

Antonialock commented 1 year ago

It seems logical to remove the taxon constraint. If the strusture has been hi-jacked my the pathogen, then it's not a "normal structure" of the host any longer - and we only annotate "healthy" processes in GO? So once it is hijacked it "belongs" to the pathogen, and those human proteins are involved in "disruption of cellular component of another organism" (even if the structure originated from the host?). Per haps this structure is even a "GO:0044217 other organism part"?

image

genegodbold commented 1 year ago

FWIW, I wouldn't give ownership to either one of them completely. I believe it is a cellular entity that has been hijacked/stolen by the pathogen. The parasite is forcibly borrowing it, but what it's borrowing is the host's lipids and enzymatic machinery. I suppose that, in some cases, it could conceivably be generated de novo (still from host cellular components) but I don't think this is the normal case. Viruses repurpose the endoplasmic reticulum and the Golgi frequently to make their viral capsid factories (SARS-CoV-2 does this). Bacteria and protozoan beasties (as T. gondii above) do this as well. [I give bacteria more credit (in my silly head) because they can't be expected to know eukaryotic biology like another eukaryote.]

Please let us not apply the "disruption of cellular component of another organism" to anything from the host, those are intended (I hope) only for symbiont/parasite proteins. There are too many symbiont terms in GO/Amigo that are contaminated with host proteins already.