1 2 aim

#apaperaday: Systemic delivery of full-length dystrophin in Duchenne muscular dystrophy mice

In today’s #apaperaday, Prof. Aartsma-Rus reads and comments on the paper titled: Systemic delivery of full-length dystrophin in Duchenne muscular dystrophy mice

Today’s pick is from Nature Communications by Zhou et al on the delivery of a full-length dystrophin in mice using AAV and the intein system to split the dystrophin code over 3 (!) different vectors. DOI: 10.1038/s41467-024-50569-6

Adeno-associated viruses (AAV) are the vector of choice for delivery to skeletal muscle and heart, but their capacity does not allow putting in the full protein code for dystrophin. Thus, current AAV gene therapy trials and the approved gene therapy for Duchenne use micro-dystrophin.

The micro-dystrophins are only 1/3 of the full-length dystrophin and are expected to be only partially functional. To compare, Becker patients have only partially functional dystrophins, and their dystrophins are usually 2/3 of full length or larger.

Authors argue that restoring a full-length dystrophin would be better as these are more functional than micro-dystrophin. BUT, as each AAV vector can contain only one third of the dystrophin code, this would mean you would have to split the code into 3 and then reassemble.

This reassembling is the difficult part as you need to have each piece within the same muscle AND the cell would need to know how to recombine the 3 pieces of dystrophin. Authors use the intein system. Inteins are small peptides that self-assemble and then induce transsplicing.

Authors used 2 different inteins, as you need to assemble part 1 to part 2 and part 2 to part 3 to get a full-length dystrophin: one type of intein to connect 1 to 2, and one to connect 2 to 3.

Authors first tested this in HEK cells with plasmids, which revealed it is possible to get full-length dystrophins this way but only at low levels. Optimization was needed, e.g., the location of where to split the dystrophin and the ratio of the different parts.

Authors tried other inteins, but that made assembly worse. After optimization, they packaged the 3 dystrophin pieces into AAV vectors. They used an optimized AAV, MYOAAV4, that homes better to skeletal muscle. They used a total dose of 2.10^14 viral genomes/kg in a 2:1:1 ratio.

They tested this in mdx4cv mice, showing dystrophin restoration after treatment. Western blot showed it was full-length dystrophin, and immune fluorescence analysis showed it was located at the sarcolemma and recruited components of the associated glycoprotein complex and nNOS.

This treatment also reduced creatine kinase levels in blood (a signal for less muscle damage), improved muscle torque, reduced necrosis, and reduced fiber size variability (all signals of pathology) in skeletal muscle. However, there was less dystrophin produced in the diaphragm.

Authors argue that perhaps the promoter was less effective in the diaphragm (it is also possible the MYOAAV was less able to infect the diaphragm). They optimized the promoter and this did improve the expression of dystrophin in the diaphragm to some extent.

Authors quantified dystrophin levels in skeletal muscle, diaphragm, and heart using human protein lysates from skeletal muscle as a reference. I do not understand the reasoning, as dystrophin levels may vary between species. They definitely vary between muscle types!

Authors state they had no access to control human heart & diaphragm protein, so they used skeletal muscle as a reference. However, we know dystrophin is expressed ~10x more in the heart than in skeletal muscle, so this is not good for quantification purposes. See this paper DMD transcript imbalance determines dystrophin levels.

Authors then compare micro-dystrophin and their full-length dystrophin in the mdx4cv mouse using 2 micro-dystrophins (based on what is used in trials but not identical). They use the same promoter for each and MyoAAV4 for delivery.

This showed micro-dystrophins are expressed at 2.5-3.5 times higher levels than normal dystrophin in controls, while full-length dystrophin was expressed at ~0.5 times. For CK, muscle function & histology, authors saw no difference between micro and full-length dystrophin in mice.

Some notes: this is in mice. Furthermore, the full-length dystrophin was expressed at lower levels. Also, authors show that signaling pathway changes were only restored for the full-length dystrophin. Also, there were no differences for the tests authors did.

I suggest treadmill running and eccentric exercise analyses to see whether there truly is no difference between the micro-dystrophins and the full-length dystrophin (I anticipate micro-dystrophins will not fully protect against treadmill running & eccentric exercise-induced damage).

Authors discuss that there was a lot of optimization involved in getting the 3-part intein split reassembly of full-length dystrophin to work at the current levels and that more optimization is needed to improve dystrophin levels further.

Authors share my worry that restoring full-length dystrophin risks an immune response to dystrophin from patients. Micro-dystrophins and Becker-type dystrophins, as restored after exon skipping, generally will not contain new epitopes and therefore there is tolerance.

However, we know that when we treat patients who lack the first part of the micro-dystrophin due to a deletion, there is a risk of an autoimmune response to all the tissues expressing micro-dystrophin. When restoring full-length dystrophins, this risk exists for all patients.

Authors outline that using immune suppression may help tolerate patients. However, the side effects of severe immune suppression should not be underestimated (especially if long-term suppression would be required).

An alternative approach may be to make the dystrophin less immunogenic by making small adaptations to the protein that do not affect function. This is something Jeffrey Chamberlain is currently attempting to make the micro-dystrophins less immunogenic.

For the current paper, the ability to restore full-length dystrophin with the intein system is conceptually nice, but I do not know whether it will be clinically translatable.