Lightnews — Scholar-powered news

Annemieke Aartsma-Rus @oligogirl.bsky.social · 2h

Checking variant databases is important as well (with the earlier mentioned caveat) and everyone should be aware that females can have dystrophinopathy as well and that family members of a patient can be a carrier. I appreciate the authors shared this information and the learnings!

Annemieke Aartsma-Rus @oligogirl.bsky.social · 3h

This is especially important for duplications (as they may not be within the gene). Segregation analysis within the family can give insight in pathogenicity (if a healthy grandfather has the same variant, likely it is benign). Physical examination of the case is important to assess muscle problems

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 3h

For some variants there are reports ranging from Duchenne to healthy and then it is difficult to classify the variant for a new case who is young. Authors conclude that when copy number variants are found in the DMD gene they have to first be validated by another technique.

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 3h

Also there is a brain co-morbidity in a subset of patients & they may be analyzed due to developmental delay, which in fact is part of the dystrophinopathy sprectrum. Authors stress the need for good reporting in databases, where sometimes very little clinical information can be found for a variant

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 3h

Authors discuss that with more genetic analyses done now at exome and genome wide levels incidental findings will happen more and more. Usually these analyses are done when children are very young, and muscular dystrophy symptoms may not yet be apparent.

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 3h

In one female 2 duplications were found, but it was not clear if they are on the same chromosome or not. She had muscle symptoms, so it is possible they are on separate chromosomes and it is also possible that a partial deletion happened to overlap with the duplication

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 3h

Some were not located in the DMD gene, while others were likely benign (e.g. exon 1-x --> still a functional protein can be produced probably). For a duplication of exon 45-51 an asymptomatic uncle was found.

1 1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 3h

There was also a case of a female dystrophinopathy with symptoms. Finally for some incidental findings, the variants were found on top of other syndromes (e.g. Angelman syndrome and female carrier of Duchenne variant). Authors could confirm the location of 11/13 duplication variants.

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 3h

Some of the variants indeed caused Duchenne or Becker and this was confirmed when patients were clinically evaluated. For some however, no symptoms were (yet) present. Interestingly for 2 cases, healthy family members with the same variant were discovered: deletions of exon 2-9 and 50-51.

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 3h

One in three variants are de novo, the rest is inherited. Authors here report on 32 copy number variations in the DMD gene that they detected as incidental findings: 19 deletions and 13 duplication variants. Validation was done with MLPA analysis which confirmed the variants in most cases.

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 3h

Deletions in the DMD gene can cause Duchenne (usually out-of-frame) or Becker (usually in-frame), female dystrophinopathy (for a subset of mutation carriers) and occur in healthy individuals. For duplications, the same applies, & the duplicated region is not per se within the gene.

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 3h

When individuals present with symptoms and it is not clear what the cause is, genome analysis is often done to see if a genetic cause can be identified. During these analyses sometimes also unexpected discoveries are done, e.g. deletions or duplications in the dystrophin encoding DMD gene.

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 3h

#apaperaday Today's pick is from @worldmusclesociety.org journal Neuromuscular Disorders by LUMC colleagues Ginjaar et al on incidental findings of deletions and duplications in the DMD gene. Yuzu took a nap so no confetti making.DOI: 10.1016/j.nmd.2025.106219

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 1d

I also do not like how authors used predictive software to see whether their model is working or not, without any experimental validation. e.g. We know Splice AI can be wrong - it is the best tool we have, but it is far from perfect.

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 1d

I like the systematic approach with which authors started. However, they did this for 1 exon in 1 transcript and I am not sure this can then be extrapolated across other exons (also since different splicing factors are expressed in different cells).

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 1d

Also there is no validation of the tool to see whether ASOs predicted to increase skipping or inclusion actually do this. Authors discuss that more work is needed as they now only focused on exons, and introns also play a role in splicing. Indeed.

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 1d

They made DANGO, a tool to predict whether an ASO is likely to induce exon skipping or inclusion, which can be used to help design ASOs. Note that this tool does not take into account that the ASO needs to be able to bind to transcripts better than to their counterparts.

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 1d

They studied 18.5 thousand exons, where the splice enhancing regions were usually short, on average 8.5 nucleotides. Shorter exons were enriched for enhancer motifs. They also used this prediction to design ASOs and predict the effect (again no validation at all with experiments).

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 1d

They studied 18.5 thousand exons, where the splice enhancing regions were usually short, on average 8.5 nucleotides. Shorter exons were enriched for enhancer motifs. They also used this prediction to design ASOs and predict the effect (again no validation at all with experiments).

Annemieke Aartsma-Rus @oligogirl.bsky.social · 1d

Finally they fed their FAS exon 6 dataset to a deep learning system, and then used that to make a model to predict the effect of substitutions and deletions, using splice predictive software to test how well the model worked (so not functional validation).

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 1d

Authors then studied other transmembrane domain coding exons, seeing that they are often pyrimidine rich (as they code for hydrophobic amino acids) and that these exons have a lower likelihood to be included. Authors argue that deletion analysis can identify exon inclusion domains.

2

Annemieke Aartsma-Rus @oligogirl.bsky.social · 1d

Authors argued that the polypyrimidine region present in FAS exon 6 could be a potential binding site for U2 snRNP and using in vitro transcribed RNA they showed this. They speculate that U2 binding in exons may be a way to induce exon skipping in normal exons (no evidence shown though).

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 1d

When authors added a substitution to generate a splice acceptor site, the micro-exon was included efficiently. Authors overexpressed SRRM4 protein, which is involved in micro-exon inclusion and could increase the level of inclusion, while the wild type exon was skipped with SRRM4 increase

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 1d

Authors systematically made the exon shorter, seeing that below 50 nt and 30 nt inclusion became less or a lot less, respectively. However, for one region, inclusion was efficient. This region had a polypyrimidine stretch and a branchpoint but no splice site.

1

Annemieke Aartsma-Rus @oligogirl.bsky.social · 1d

The exception was for insertions near the end of the exon - there it led to exon skipping. Purine rich insertions also increased exon inclusion (purine rich sequences are often found in splicing enhancers so this makes sense).

1