In today’s #apaperaday, Prof. Aartsma-Rus reads and comments on the paper titled: Transiently expressed CRISPR/Cas9 induces wild-type dystrophin in vitro in DMD patient myoblasts carrying duplications.
Today’s pick is from Scientific Reports by Pini et al (from Muntoni’s group) on genome editing for duplication mutations in Duchenne muscular dystrophy. Proof of concept work in cultured cells. Doi 10.1038/s41598-022-0767-w
Duchenne is caused by lack of dystrophin protein ~10% of patients has a duplication of multiple exons. Approved Duchenne therapies do not apply (stop codon readthrough & exon skipping). Microdystrophin gene therapy applies. However, the function of microdystrophin is yet unknown.
NB a clinical trial for exon 45, 51 and 53 exon skipping for single exon duplications is ongoing: as is a clinical trial for single exon 2 duplications. Here the focus is on multiple exon duplications.
For multiple exon duplications genome editing would allow a single guide RNA approach. This cuts at 2 locations allowing removal of the duplicated area. The feasibility of this was previously shown by Dwi U. Kemaladewi, PhD et al for a duplication of exon 18-30.
Authors reason permanent expression of cas9 is undesirable due to the risk of irreversible, unwanted cuts in other areas of the gene by cas9 going rogue. They propose a transient cas9 expression system & optimize this in primary cells from a patient with an exon 3-16 duplication.
An exon 3-16 duplication is in-frame, and therefore these cells produce a 499 kDa partially functional dystrophin. This means that authors can study whether the normal dystrophin is restored on western blot (427 kDa) and whether too large dystrophin is reduced.
Authors target intron 9 as this is a region that has little polymorphisms and it is within the duplication hotspot of the gene (exon 2-20). Note that it does not matter where within the duplication the guide is designed as it will always restore normal dystrophin.
This is a very nice aspect of genome editing for multiexon duplications and also means that one can design a single guide RNA to be applicable to multiple Duchenne patients with a duplication of exons. Authors first confirm their guide RNA works using lentiviral vectors.
Using this ‘permanent expression system’ they observed ~29% editing and ~50% of normal dystrophin production. So the guide RNA works, and now authors focus on delivering the cas9 transiently using different transfection reagent to deliver the 4.7 kb plasmid.
The transfection efficiency is very low (everyone trying to transfect plasmids to primary myoblast cultures will feel the author’s pain: this is very challenging). Authors then switch to electroporation and using this method achieve an efficiency of ~30%.
Due to a GFP tag in the plasmid they could do FACS sorting of transfected cells. However, when they then cultured the treated cells they went into senescence (again primary myoblast experts will recognize this challenge: primary myoblasts have limited proliferative capacity).
Authors immortalized their myoblast culture. Now electroporation is slightly less effective, but still results in ~20% editing and production of ~45% of wild type dystrophin. Authors also notice that some cells took up more plasmid – here dystrophin restoration is higher.
Authors conclude that they show proof of concept of the transient genome editing approach. However, what is not clear is how long lasting the cas9 expression is (cannot discover this in differentiated myotubes as the cultures have a limited shelf life).
Authors discuss that their electroporation approach is not achievable in vivo. AAV has been used so far but this provides longtime expression of cas9 and with turnover of muscle fibers, edited nuclei will be lost. AAV is unable to efficiently reach satellite cells.
With turnover there will be a dilution of edited nuclei while it is not possible to retreat due to an immunity against AAV. Provided efficient editing, my question is how much muscle turnover there would be for duplications since normal dystrophin would be produced.
The dilution in my opinion is more a problem when Becker-type dystrophins are produced, again provided dystrophin levels are sufficient and editing is efficient (as yet we do not know the level of ‘sufficient’, probably somewhere between 20 and 50%).
Back to publication: authors suggest nanoparticles to deliver cas9 & guideRNA plasmids to muscle. Currently nanoparticles do not allow systemic delivery to skeletal muscle. Hopefully doable in the future. For repeated treatments antibodies against cas9 may still be an issue.
The authors end by stressing that for genome editing a ‘hit and run’ strategy is preferred over long term expression of cas9. I agree with this – when delivering cas9 systemically long term expression holds too many risks for aberrant edits.
All in all a nice paper but very early stage work. The fact that ‘it works’ in cell cultures does not mean it can work in patients. More work is needed (I do realize that my thread often end like this).
Pictures by Annemieke, used with permission.