#apaperaday: Integrated genomic, proteomic and cognitive assessment in Duchenne Muscular Dystrophy suggest astrocyte centric pathology

In today’s #apaperaday, Prof. Aartsma-Rus reads and comments on the paper titled:  Integrated genomic, proteomic and cognitive assessment in Duchenne Muscular Dystrophy suggest astrocyte centric pathology

Today’s pick is from Heliyon by Wijekoon et al on genomic, proteomic and cognitive assessment of a cohort of 25 Duchenne patients from Sri Lanka DOI: 10.1016/j.heliyon.2023.e18530

Authors introduce that it is now recognized that Duchenne also afflicts the brain and that many patients have cognitive problems, learning difficulties and behavioral problems. This makes sense as different dystrophins are produced in the brain.

There are 2 full length isoforms expressed in brain (similar to the muscle isoform expect for the very first beginning), a medium sized isoform Dp140 and a very short isoform Dp71. The promoters for Dp140 and Dp71 are located in intron 44 and intron 62.

Based on where Duchenne patients have their mutation, only the full length isoform is missing OR also shorter isoforms Dp140 or Dp140 & Dp71. Note that Becker patients produce partially functional full length dystrophins, but they can miss one or more shorter isoforms!

Back to Duchenne, authors outline that most of the studies into Duchenne cognition have been done on Caucasian patients. Here they have analyzed a cohort of 25 patients from Sri Lanka with cognitive tests and also have performed serum analysis for biomarkers.

All patients have a confirmed genetic diagnosis using MLPA. Authors used this to check which isoforms were missing and this is where things went wrong for assessing if patients could produce Dp140 (note that this happens often because Dp140 transcript is tricky).

Some background: the promotor for Dp140 (so the signal to start producing RNA) is located in intron 44 (so between exon 44 and exon 45). So the mRNA transcript for Dp140 is Dp140promotor, exon 45,46, 47, 48, 49, 50, 51, 52 etc until exon 79.

However, the location where the protein translation starts is in exon 51. So what you need for making Dp140 is the location in intron 44 to start making the RNA and exon 51-79. It is not clear whether missing exon 45-50 will impact Dp140 production.

Authors did not take this latter caveat into account. They also did not take into account that when deletions involve intron 44 (i.e. ending with exon 44 or starting with exon 45) they may delete the promotor for Dp140, but they may also not delete it.

So for these patients the ability for Dp140 production is unknown (unless you do more detailed DNA or RNA analysis – which authors did not do, they used MLPA). Instead authors say patients with deletions until exon 44 CAN make Dp140, while those starting with exon 45 cannot.

This means that their group division (group 1 all isoforms lost, group 2 can only produce Dp71 and group 3, can produce Dp140 and Dp71) is not correct. Authors found that 40% of the patients in their cohort had developmental delay (similar to what is found in Caucasians).

However, their downstream analyses where they try to correlate groups with cognitive impairments are not relevant as they misidentified who might be able to produce Dp140 – or Dp71 for that matter (a deletion of exon 61-62 may abolish Dp71 promotor, authors did not consider this)

I’ll not summarize what authors found here as it is not correctly done anyway. Will point out however that authors made many correlations and used p 0.05 as the threshold for significance. Note that if you make 20 comparisons you will find one significant by chance.

This is why multiple testing correction is needed – which authors did not use. Authors also looked into serum and found muscle damage markers as expected. They also performed a specific analysis to look into cognition markers and found fibronectin 1 and Siglec-3.

Fibronectin 1 levels correlated with better cognitive performance, which is strange as this is a marker of neuronal damage. Authors speculate in the discussion that Siglec-3 is a protein also found as a marker for Alzheimer and that this protein is involved in folding proteins.

They further speculate that it may be involved in facilitating phagocytosis of misfolded Dp140…this is very farfetched as I am not aware of Dp140 protein accumulation and usually what we see in muscle is that dystrophins that do not contain their C-terminal domain are instable.

Authors insist on an overlap between Duchenne and Alzheimer pathology which again is very farfetched as the brain pathology in Duchenne does not seem to be progressive and also the type of pathology does not resemble Alzheimer (learning difficulties vs short term memory loss).

Authors finally speculate about the role of astrocytes in Duchenne brain pathology. There may be a role, but I do not think authors can conclude this from the serum analysis of 25 patients. Authors do stress that they had a cohort of a limited size, which is a limitation.

They outline that more work is needed. Indeed, however, I think they first need to correctly put the patients in the right groups and repeat the analysis and comparison as this is currently not properly done.