Problems with intronic_splicing_enhancer and exonic_splicing_enhancer

[davidwsant]

Brought to my attention by Colin Logie

What is the SO term name and accession? intronic_splicing_silencer (SO:0002056) exonic_splicing_silencer (SO:002058) Both are children of silencer (SO:0000625) Also relevant, splice_enhancer (SO:0000344)

Describe what you would like to change. There are multiple issues with the current ‘cis-regulatory module’ (CRM, SO:0000727) child term ‘silencer’ (SO:0000625) and its current two child terms SO:0002056 and SO:0002058. Firstly, in the field, a transcription silencer has been equated to a ‘negatively acting ‘enhancer’, a CRM that is bound by DNA-binding repressors of transcription, such as those that nucleate the formation of heterochromatin to silence the transcription of a gene.

Second, the potential confusion between SO:0002056 and 2058 with RNA splicing control region terms that are in the SO:0001056-SO:0000344-SO:0000683-SO:0000320 branch.

Thirdly, the current molecular scenario captured by the SO:0002056 and SO:0002057 terms that may have to be accommodated, whereby unspliced mRNA binding proteins would influence DNA transcription by a DNA-template-dependent RNA polymerase. Notably, to the best of my knowledge, as of 2020, no, or very scant peer-reviewed experimental data-based literature has documented such a scenario.

Fourth, the converse scenario from 2056 and 2058, namely whereby a DNA-bound transcription factor that promotes DNA looping, influences splice site selection has been reported (PMID: 29199022). It may therefore be good to include that in SO.

Specific issues and possible solutions

1. The definition of SO:0000625 silencer is correct. The issue are the current children SO:0002056 and SO:0002058, that are very special cases of RNA sequence-dependent DNA transcription regulation.

2. The ancestor terms’ (SO:0000625 and SO:0000727) definitions include transcription factors, which are DNA binding proteins that influence the rate, frequency and/or efficiency of RNA biosynthesis by RNA polymerases, whilst the two child terms are about RNA elements and RNA binding proteins that influence RNA splicing by the spliceosome if they were to be referred to as splicing factors (this is explicit in the literature reference PMID: 23241926 . This can be solved by moving the ontology branch consisting of the two current children of SO:0000625 to another parent term that would represent CRMs that act as RNA elements that are bound by RNA binding proteins to influence transcription, contrary to the enhancer (SO:0000165) and the locus control region (SO:which are CRMs that act at the DNA sequence element level by virtue of being bound by DNA binding proteins.

3. Note that the SO:0000344 branch is correct as a child of splicing_regulatory_region (SO:0001056). But, the parent of SO:0001056 should not be transcription_regulatory_region (SO:0001679) as these are not transcription regulatory elements. It is therefore misplaced. SO:0001056 should be a child of SO:000831.

4. There is a parental CRM term that encompasses RNA and DNA cis-regulatory modules; those that respectively act at the RNA level and those that act at the DNA level. It is SO:0000831. A new child for this term, next to transcription (SO:, translation and recombination control regions, would be a splicing control region (currently splicing_regulatory_region (SO:0001056)).

5. The converse scenario from 2056 and 2058, namely whereby a DNA-bound transcription factor that promotes DNA looping, influences splice site selection has been reported (PMID: 29199022). It may therefore be good to include that in SO. So a spicing regulation branch that depends on DNA-bound proteins that influence mRNA splicing.

Relevant Publications PMID:23241926 speaks about silencers on RNA affecting the splicing of RNA as opposed to silencers on DNA that affect transcription.

[davidwsant]

Hi Colin,

I think this can be fixed without changing that many terms. Below is my proposal:

Move splicing_regulatory_region: id: SO:0001056; name: splicing_regulatory_region; def: "A regulatory_region that modulates splicing." Change to is_a: SO:0000831 ! gene_member_region

Create a new child term: name: splice_silencer; def: "An RNA splicing regulatory element that functions to recruit trans-acting splicing factors suppress splicing." [PMID:23241926, SO:ke] is_a: SO:0001056 ! splicing_regulatory_region

Move the two misplaced terms: id: SO:0002058; name: exonic_splicing_silencer; def: "An exonic splicing regulatory element that functions to recruit trans acting splicing factors suppress the transcription of the gene or genes they control." [PMID:23241926, SO:ke] is_a: New_Term ! splice_silencer (obviously I will add the ID once it is created) is_a: SO:0000852 ! exon_region

id: SO:0002056; name: intronic_splicing_silencer; def: "An intronic  splicing regulatory element that functions to recruit trans-acting splicing factors suppress the transcription of the gene or genes they control." [PMID:23241926, SO:ke] is_a: New_Term ! splice_silencer relationship: part_of SO:0000188 ! intron

I think this will cover the issues. Colin, do you agree? @keilbeck, do you agree with these changes?

Thanks,

Dave

[colinlog]

I concur Dave, except for the literature reference for SO:0002056 and SO:0002058 as PMID:23241926 does not mention transcription regulation at all. It is a review about splicing control and appears to me to be about the terms in the SO:0001056-SO:0000344-SO:0000683-SO:0000320 branch. Quote: "These SREs are conventionally classified as exonic splicing enhancers (ESEs) or silencers (ESSs), and intronic splicing enhancers (ISEs) or silencers (ISSs)." I do not know literature about non-viral eukaryotic RNA sequences and the proteins they would recruit to act as transcription regulator activity related to that of the sequence-specific DNA-binding Transcription Factors (GO:0003700). A viral example is TAR (GO:1990970) and that is activating rather than silencing transcription.

Small thingie: "... to recruit trans-acting splicing factors suppress the transcription of the gene or genes they control ..." should read "... and suppress ...", shouldn't it?

Having though more about this, I propose to erase SO:0002056 and SO:0002058, as well as SO:0002057 since the latter goes nowhere and appears to me to be entirely redundant with the SO:0001056-SO:0000344-SO:0000683-SO:0000320 branch for splicing regulation.

The alternative is to keep SO:0002056 and SO:0002058 for transcription repression and to create 2 more terms for transcription activation. However, in Wikipedia the TAR RNA is referred to as a microRNA. I am therefore unsure that it could be referred to as as either an intron or an exon. Hence, the 2056 -2058 branch appears to need more work, since it lacks a transcription activation sub-branch and since exons and introns may not be sufficient nor appropriate to describe the example of RNA-directed transcriptional control in eukaryotic cells that is known to me, namely the TAR-TAT system.

Best, Colin

@davidwsant @keilbeck

[davidwsant]

Hi Colin,

I see what you mean. The definitions mention transcription. However, when they were added they were requested by someone that suggested the publication. I believe he meant to have them added as terms related to splicing, not related to transcription. I think the definitions need to be updated to reflect that, along with location change in the ontology. You mentioned that they should be obsoleted, and because the definition change is significant and they would be changing locations, I think that is appropriate. I do not like to obsolete terms if at all possible, but I think this is one of the circumstances where it may be needed. I have simply shortened the definitions for these two terms to make them about splicing only and not about transcription. Could you please review these changes?

Proposed changes:

Move splicing_regulatory_region: id: SO:0001056; name: splicing_regulatory_region; def: "A regulatory_region that modulates splicing." Change to is_a: SO:0000831 ! gene_member_region

id: New_Term; name: splice_silencer; def: "An RNA splicing regulatory element that functions to recruit trans-acting splicing factors and suppress splicing." [PMID:23241926, SO:ke] synonym: "splice silencer" EXACT [] is_a: SO:0001056 ! splicing_regulatory_region

id: SO:0002058 will be obsoleted and a new ID will be generated; name: exonic_splicing_silencer def: "An exonic splicing regulatory element that functions to recruit trans-acting splicing factors." [PMID:23241926, SO:ke] synonym: "ESS" EXACT [] synonym: "exonic splicing silencer" EXACT [] is_a: New_Term ! splice_silencer is_a: SO:0000852 ! exon_region

id: SO:0002056 will be obsoleted and a new ID will be generated; name: intronic_splicing_silencer def: "An intronic  splicing regulatory element that functions to recruit trans-acting splicing factors." [PMID:23241926, SO:ke] synonym: "intronic splicing silencer" EXACT [] synonym: "ISS" EXACT [] is_a: New_Term ! splice_silencer relationship: part_of SO:0000188 ! intron

Do you think that this looks more correct?

Thanks,

Dave

@colinlog @keilbeck

[davidwsant]

I accidentally clicked "close and comment" instead of "comment". It is now re-opened.