#apaperaday: Split intein-mediated protein trans-splicing to express large dystrophins

In today’s #apaperaday, Prof. Aartsma-Rus reads and comments on the paper titled: Split intein-mediated protein trans-splicing to express large dystrophins

Today’s pick is again about intein transsplicing to restore large dystrophins in Duchenne mouse models. It is from Tasfaout et al in Nature publication was out 4 days before yesterday’s #apaperaday but first submitted a year earlier. DOI: 10.1038/s41586-024-07710-8

Authors explain that dystrophin is lacking in Duchenne patients and that gene therapy can only deliver micro-dystrophins that contain 33% of the full length dystrophin. It is assumed that larger dystrophins would be more functional and therefore authors looked into inteins.

If you read yesterday’s thread you will remember that inteins are peptides that induce transsplicing. This leaves a ‘scar’ in the protein of a few amino acids, but it is assumed that dystrophin can handle this without losing function when this occurs in the rod domain.

Authors here first tested a library of 23 different intein pairs with a GFP construct to see which ones would work most efficiently, leaving the smallest footprint (scar) AND using 2 pairs that did not cross react to avoid linking part 1 to part 3 instead of 1 to 2 to 3.

Authors first made a 2-part midi-dystrophin that contains most functional domains. The graph below nicely outlines the difference between the midi-dystrophin and the micro-dystrophin used in trials (see image) In HEK cells, the plasmids were able to combine into midi-dystrophin.

When this worked well, authors attempted the full-length dystrophin, which also worked. Then, authors put the different pieces into an AAV6 vector and treated mice with a local injection into the tibialis anterior muscle. This resulted in midi and full-length dystrophin production.

Authors also tested the combination of part 1 and part 3, but this did not result in a combined protein, as expected since these inteins should not induce trans-splicing (nice control though!). Authors then used AAV6 for systemic treatment using midi, full-length & micro-dystrophin.

For the micro-dystrophin, they used AAV9, as that is also used in trials. AAV6 did not lead to good infection, so authors moved to AAVmyo, just like yesterday’s authors. They used a total dose of 2.10^13 and 2.10^14 viral genomes per kg, and for micro-dystrophin, they used AAV9.

Authors claim that the midi-dystrophin does better than the micro-dystrophin. However, I am not that convinced (and saw reviewers were wondering as well in the reviewer reports). Indeed, mdx mice where micro-dystrophin was delivered with AAV9 do worse than AAVmyo midi-dystrophin.

However, when AAVmyo was used to deliver micro-dystrophin, most of the results are the same. Authors then used the same treatment groups in 17-month-old mice, which have more severe pathology and also cardiac problems. Here, the midi-dystrophin did do better than micro-dystrophin.

The specific force was improved in skeletal muscle, and cardiac function and diaphragm force were better for midi- than micro-dystrophin. What does this mean? Bear in mind all of this is in mice. Yesterday’s paper found no difference between micro-dystrophin & full-length IN MICE.

If the same holds true for humans, we do not know. Yesterday’s paper did not look into eccentric contractions. Today’s authors DID and they saw (slightly) more protection by midi-dystrophins. I would still be interested to see the difference between groups after treadmill running.

Authors discuss that more work is needed to optimize the approach, e.g., by using different ratios of the pieces (yesterday’s paper did this and saw that it would increase dystrophin restoration). Authors also stress that work is needed to study the safety of the inteins.

Will an immune response occur to the inteins, or the scars they leave in the protein (may be neo-epitopes for the immune system)? Like yesterday, also today authors worry that restoring full-length dystrophin may be immunogenic, and therefore a midi-dystrophin may be preferred.

It was interesting to read the papers back-to-back and to see how there were similar findings (in mice, micro-dystrophin does quite well) and similar approaches (for inteins to work, we need to go beyond currently used AAVs into the optimized myoAAV/AAVmyos).