NTR: GO:0035064 methylated histone residue binding

[gocentral]

Dear GO editors, according to PubMed=23948251, CHD5 specifically binds H3K27me3 modified histones in chromatin through its chromodomains. On the other hand, PubMed=23318260, finds that the PHD domains may mediate specific binding to unmethylated lysine 4 of histone 3. I could find GO:0035064 methylated histone residue binding but not child terms that would allow to better describe the activity of CHD5. Would it be possible to create such child terms or is it on purpose that GO:0035064 methylated histone residue binding has no child term? Thanks for your help.

Reported by: lbreuza

Original Ticket: geneontology/ontology-requests/10769

[gocentral]

If the methylated forms are in PRO then we can trivially add the corresponding binding terms

Original comment by: cmungall

[gocentral]

• assigned_to: David Hill

Original comment by: ukemi

[gocentral]

Do you know if it binds the methylated residue itself? The parent term specifically refers to binding the methylated residue. If this is what you would like, I will add:

H3K27me3 methylated histone residue binding

Def----Interacting selectively and non-covalently with a lysine 27-trimethylated residue on histone 3. Histones are any of a group of water-soluble proteins found in association with the DNA of plant and animal chromosomes.

Please let me know by commenting in this SF item.

Original comment by: ukemi

[gocentral]

Dear David,

I am now a little confused by your question since I have no definitive proof that it binds the modified residue but here is what I could find in the above mentioned reference: "Peptide pull-downs were performed on mouse brain nuclear lysates using biotin-tagged modified histone H3 peptides. Western blots were performed for CHD5, EZH2, and HP1γ as indicated." From there I would conclude that it binds preferentially histone H3 peptides containing modified Lysine residue. More generally, there are regions of chromatin where histones are differentially modified and the binding of proteins to chromatin is affected by those modifications. This is what I would like to capture. Otherwise, your proposition sounds good, but maybe I am missing something.

Thanks for your help. Lionel

Original comment by: lbreuza

[gocentral]

• status: open --> closed-accepted

Original comment by: ukemi

[gocentral]

Hi Lionel,

We discussed this today on an editors' call and decided that we would relax these terms so that the binding did not have to be to the residue, but rather to the modified histone. We felt that most experiments would show that rather than residue binding and the purpose of these terms should be to describe binding to the modified protein. Added: [Term] +id: GO:0061628 +name: H3K27me3 modified histone binding +namespace: molecular_function +def: "Interacting selectively and non-covalently with a histone H3 in which the lysine residue at position 27 has been modified by trimethylation." [GOC:dph, PMID:23948251] +synonym: "H3-K27me3 modified histone binding" EXACT [GOC:dph] +is_a: GO:0035064 ! methylated histone binding +created_by: dph +creation_date: 2014-05-14T15:36:53Z + +[Term] id: GO:0065001 name: specification of axis polarity namespace: biological_process

Original comment by: ukemi