NTR: Fibroblast growth factor signalling pathway (and the positive regulation term)

[milarolo]

Hi, 1- New term request: Fibroblast growth factor signalling pathway 2- New term request: Positive regulation of fibroblast growth factor signalling pathway The request is supported by Fig. 3C, 4A, 6 in PMID: 16774925. SMOC2 positively regulates fibroblast growth factor (FGF2).

Child term (of term 1): GO:0008543: Fibroblast growth factor receptor signalling pathway

The same paper shows that SMOC2 positively regulates VEGFA. There is already a term for that: GO:1900748: Positive regulation of vascular endothelial growth factor signalling pathway and also for the pathway: GO:0038084: Vascular endothelial growth factor signalling pathway.

Thanks, Mila

--

[pgaudet]

Hi @milarolo We already have
GO:0008543 fibroblast growth factor receptor signaling pathway GO:0040036 regulation of fibroblast growth factor receptor signaling pathway and the positive and negative regulation terms.

We should add synonyms.

Thanks, Pascale

[milarolo]

Thank you very much for your reply. I saw those two terms and I was going to use them. But the paper doesn't say if the fibroblast growth factor (bFGF) only binds to the FGF receptor or if it binds to other receptors as well. Since GO has the GO:0038084: Vascular endothelial growth factor signalling pathway term, I thought it would be better to have a term without the "receptor" word in it.

Best, Mila

[pgaudet]

Glancing at the paper I don't see any assay for a specific signaling pathway - Fig. 3C looks at cell proliferation

• Fig 4A looks at cell migration
• Fig 6 looks at angiogenesis

If you use Noctua to annotate you can annotate SMOC2 to 'acts upstream of' these 3 processes. The paper does not seem to support any stronger statement about SMOC2's specific role in the signaling process or its regulation.

@vanaukenk Do you agree?

Thanks, Pascale

[milarolo]

Thank you for your comments. I agree that there are no experiments for specific signalling pathways. However VEGFA and FGF2 are growth factors that are part of signalling pathways involved in vascular development and new blood vessel formation. The assays in those figures show a clear positive regulation of VEGFA and FGF2 by SMOC2 when they are looking at their effects in cell proliferation, cell migration (which are main events that happen during the process of angiogenesis) and angiogenesis. Therefore we can infer that SMOC2 positively regulates VEGFA and FGF2 signalling pathways.

Mila

[ukemi]

This is an interesting example. If we feel that there is enough evidence for pathways based on the things the pathways control, we would still annotate to the pathways that are based on the activity of the receptor. Unless we have recently changed our policy, signaling pathways are differentiated by the receptor and it's downstream causal targets. Differentiating the pathways based on ligands was thought to be too problematic since a single ligand may bind to many receptors.

This might be a good candidate for a GO_CAM model.

[ukemi]

Either way, I don't think this is an ontology issue. We wouldn't want to make a new term for any pathway downstream of the ligand. The question that remains here is the annotation.

[vanaukenk]

We have these two terms in the ontology, each with specific annotation comments:

GO:0048010 vascular endothelial growth factor receptor signaling pathway (65 experimental annotations) Definition Any series of molecular signals initiated by the binding of an extracellular ligand to a vascular endothelial growth factor receptor (VEGFR) located on the surface of the receiving cell, and ending with regulation of a downstream cellular process, e.g. transcription. Source: GOC:ceb, PR:000001971, GOC:signaling Comment In GO, a gene product with 'vascular endothelial growth factor-activated receptor activity ; GO:0005021' necessarily binds VEGF to transduce a signal. In contrast, the VEGFR refers to PR:000001971. To represent cross-talk between ligands and receptors, signaling pathways in GO are starting to be named after the receptor and/or the signal. GO:0048010 is for annotation of any pathway in which a ligand (VEGF or an alternative growth factor) binds and activates a VEGFR (PR:000001971). For annotation of signaling pathways where a VEGF binds to a cell surface receptor (VEGFR, PDGFR etc.), consider 'vascular endothelial growth factor signaling pathway ; GO:0038084'.

GO:0038084 vascular endothelial growth factor signaling pathway (8 experimental annotations) Definition A series of molecular signals initiated by the binding of a vascular endothelial growth factor (VEGF) to a receptor on the surface of the target cell, and ending with regulation of a downstream cellular process, e.g. transcription. Source: PMID:17470632, PR:000017284, PR:000017285, PR:000003096, PR:000007520, GOC:signaling Comment In GO, a gene product with 'vascular endothelial growth factor-activated receptor activity ; GO:0005021' necessarily binds VEGF to transduce a signal. To represent cross-talk between ligands and receptors, signaling pathways in GO are starting to be named after the receptor and/or the signal. GO:0038084 is for annotation of any pathway in which the ligand VEGF binds and activates any cell surface receptor (VEGFR, PDGFR etc.). For annotation of signaling pathways where a VEGFR binds one of its physiological ligands (VEGF or an alternative growth factor), consider 'vascular endothelial growth factor receptor signaling pathway ; GO:0048010'.

The latter term really does seem to be from the point of view of a VEGF ligand, while the former from the point of view of a specific receptor.

We don't mirror this representation with the FGF receptor signaling pathway (or, for example, the TGFbeta receptor signaling pathway), so I can see why it might be confusing.

I think we've also been grappling with this issue a bit wrt the Wnt signaling pathway and norrin signaling, see #7836

Should we propose to obsolete GO:0038084 and then possibly also revise the definition of GO:0048010?

[pgaudet]

Hi @vanaukenk @milarolo A few more thoughts about this:

• For this specific case though, I agree that obsoleting GO:0038084 seems appropriate, since the paper used to support that term is really looking at PDGFR signaling (GO:0048008).
• As one of our ontology principles, we should avoid creating similar terms that can lead to inconsistent annotations.
• As I mentioned earlier, with respect to the paper that @milarolo wants to annotate, no specific pathway is investigated, only downstream effects.
• Generally I don't think we are ready to have a strict rule to make terms only describing ligands or receptors; @ukemi had a useful functional definition, ie which are the downstream causal targets; see also the ERBB ticket #15661: the activation of multiple receptors by each ligand seemed to make it more logical to use the ligand rather than the receptor (it may be possible to work on that branch and define specific pathways better). We have a number of ligands in the ontology (cytokines, for example), we should check if it's possible to move these to their respective receptors but that's a big project.

What do you think ?

Pascale

[pgaudet]

See some notes about this here (with respect to the cross-talk between receptors): http://wiki.geneontology.org/index.php/Non_1:1_receptor:ligand_binding

[RLovering]

I think the problem that the multiple ligands bind 1 receptor and 1 ligand bind multiple receptors is a big issue. Off the top of my head I do not know if this is always via ligand families binding receptor families, or if a ligand can bind unrelated receptors and if a receptor can bind unrelated ligands. But considering family groups and then using AE/Noctua might be the easiest way forwards. Ruth

[deustp01]

My bystander impression is that the phenomenon of multiple ligands able to interact with multiple receptors is specific to the FGFR receptor family and its FGF ligand family. However, Mohammed Moosa at NYU is an authority on this - he worked out the structural biology for FGFs and FGFRs, and is probably knowledgeable about this is general, so you might be able to get a sense of whether this is a one time problem for FGF(R) or a more general one. Mohammed.Moosa@nyumc.org

[RLovering]

I think the interleukins have this problem too. eg statement in abstract: https://www.ncbi.nlm.nih.gov/pubmed/29736903 IL-4 exerts its activities by interacting with two specific cell surface receptors: one designated the type 1 IL-4 receptor (IL-4R); the other designated the type 2 IL-4R, a receptor complex that is also the functional receptor for IL-13. And Fig 1 in https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5897441/ which confirms 3 ligands for IL-20R type I receptor and 2 ligands for IL-20R type II receptor. I think Alex Diehl might be able to offer some advice too. @addiehl

[RLovering]

Please do not obsolete GO:0038084 vascular endothelial growth factor signaling pathway

There are 98 EXP annotations associated with the term (if you include the regulation child terms) weirdly this term also has the 5021 child term see attached screen shot, and 38190 child confirms why VEGF signaling is not a child of VEGFR signaling.

[pgaudet]

Isn't this a case where the term 'receptor transactivation' could be used ?

[milarolo]

Hi, Is there any update on this issue? Thanks, Mila

[hdrabkin]

It appears to me to still be in flux; I will not create the terms unless there is agreement

[milarolo]

Any decision taken on this issue?

[hdrabkin]

So the original post is deemed an annotation issue still?

[hdrabkin]

will make synonym and add comment for usage.