hypoblast - zebrafish vs mammals

[gocentral]

GO has a fairly standard mammalian definition of hypoblast:

GO:0090008 ! hypoblast development [DEF: "The process whose specific outcome is the progression of the hypoblast over time, from its formation to the mature structure. The hypoblast is a tissue formed from the inner cell mass that lies beneath the epiblast and gives rise to extraembryonic endoderm."]

I have open Gilbert Fig 11.32, the GO text def is consistent with the lineage diagram here

There is currently one annotation - oep, a zebrafish gene

However, in zebrafish, the term "hypoblast" has a different definition

id: ZFA:0000117
name: hypoblast
def: "The inner of the two layers of the blastoderm that forms during gastrulation and give rise to the definitive mesoderm and endoderm." [ZFIN:ZDB-PUB-961014-576]
synonym: "mesendoderm" EXACT [ZFIN:ZDB-PUB-961014-576]

The lineage is different, even though it may be analagous in terms of spatial configuration.

The lineage actually fits the embedded GO term here:

GO:0048382 ! mesendoderm development [DEF: "The process whose specific outcome is the progression of the mesendoderm over time, from its formation to the mature structure. In animal embryos, mesendoderm development gives rise to both mesoderm and endoderm tissues."]
http://www.ncbi.nlm.nih.gov/pubmed/14499649

VHOG has a status "uncertain" homology association between XAO:mesendoderm and mammalian hypoblast

Representing early development in a multi-species way is a challenge!

Reported by: cmungall

Original Ticket: geneontology/ontology-requests/9676

[gocentral]

• assigned_to: nobody --> ukemi

Original comment by: ukemi

[gocentral]

From the zebrafish perspective it looks like the GO:0090008 ! hypoblast development conflates anatomical structure development with developmental fate. Hypoblast is the lower of the 2 embryonic layers present in the blastoderm.

We could change the definition of hypoblast development to that of a purely anatomical structure
GO:0090008 ! hypoblast development [DEF: "The process whose specific outcome is the progression of the hypoblast over time, from its formation to the mature anatomical structure. The hypoblast is a tissue formed from the inner cell mass that lies beneath the epiblast."]

Then we could change the definition of mesendoderm to reflect that it is a fate specification
GO:0048382 ! mesendoderm development [DEF: "The process whose specific outcome is the progression of the mesendoderm fated tissue over time, from its formation to the mature structure. In animal embryos, mesendoderm development gives rise to both mesoderm and endoderm tissues."]
http://www.ncbi.nlm.nih.gov/pubmed/14499649

If we separate anatomical structure from tissue fate we should put notes to those using the terms that they should consider annotating gene functions on both axises of clasification.

Early embryonic development is difficult to generalize over time for example from Kimmel et all 1995. describing zebrafish development "Just as there was no blastocoele during the blastula period, there is no archenteron in the gastrula. Neither is there a blastopore; DEL cells involute at the blastoderm margin, which thus plays the role of a blastopore."

Original comment by: cerivs

[gocentral]

Thanks Ceri,

Hi Ceri,

I don't think we should base any definitions referring to anatomy in GO solely on developmental lineage. We have purposely tried to avoid lineage in the ontology because, as alluded to below, it is very difficult to create universal rules for lineage relationships. I like changing the definition of hypoblast dev to be more anatomical. I'm less comfortable with having a strictly lineage-derived relationship for the mesendoderm. I would be comfortable with having mesendoderm be a type of hypoblast that gives rise to mesoderm and endoderm because it would still fit in the primary anatomy heirarchy. Ultimately this representation will be up to the folks working on the anatomies though. What do you think?

-David

Original comment by: ukemi

[gocentral]

You can have lineage in the definition of the anatomical structure without necessarily incorporating lineage into the GO relationships.

However, GO already is full of implicit lineage. E.g heart field specification is part of heart formation.

But anyway, I think we have to get the anatomy right first..

Original comment by: cmungall

[gocentral]

Yes, there are some areas of GO where the lineage is certainly going to creep in. As you mention, heart field is one, anything having to do with fate commitment is another. But, I think we should get that lineage information from external ontologies.

One thing we really need to decide is when an anatomical structure changes from one identity to another one and is no longer the progression of the same structure. This is critical in the way we represent development in GO. So in the case of the heart field, the heart field, would be a heart. We can talk more about this next week. I have some examples in my file.

Original comment by: ukemi

[gocentral]

I am happy with just changing the definition of hypoblast to be more structure based.

I am happy to let the mesendoderm term fall wherever it should in the tree. (I am not sure what the mature structure of mesendoderm looks like and don't want to have to figure it out).

Original comment by: cerivs

[gocentral]

I am not sure what the mature structure of mesendoderm looks like and
don't want to have to figure it out).

Me neither, but as I said in my previous comment, someone is going to have to deside when mesendoderm becomes something else. Before it does then it is mature mesendoderm.

:)

Original comment by: ukemi

[gocentral]

Turns out having mesendoderm as a type of hypoblast won't work because the worm also uses mesendoderm. We could modify the def of mesendoderm to illustrate that it is defined by fate as Ceri has it below.

Original comment by: ukemi

[cmungall]

Revisiting. Thinking of this for uberon:

mesendoderm: region of tissue that gives rise to both endoderm and mesoderm

I don't think this is a germ layer - pre germ layer?

Then we can let figure out all the different species-specific hypoblast definitions, see where they fall out as @cerivs suggests

[pgaudet]

@cerivs @cmungall Is there an action here ? Otherwise I will close this.

[cerivs]

Can we change the def to

The process whose specific outcome is the progression of the hypoblast over time, from its formation to the mature structure. The hypoblast is a tissue formed from the inner cell mass that lies beneath the epiblast.
The portion of the definition defining the fate of the hypoblast is wrong for zebrafish. If the definition isn't changed we need a new hyboblast term for zebrafish with the correct developmental outcome.
https://zfin.org/zf_info/zfbook/stages/figs/fig13.html