ARTICLE IN BRIEF
Three different research groups were able to use the CRISPR/Cas9 gene editing system to permanently correct the dystrophin mutation in a mouse model of Duchenne muscular dystrophy.
The CRISPR/Cas9 gene editing system can permanently correct the dystrophin mutation in the mouse model of Duchenne muscular dystrophy (DMD), according to three independent studies that were published online December 31, 2015, in the journal Science. The papers provide proof of principle for considering clinical trials in boys with DMD.
“The advantage the CRISPR technology has is that it creates an irreversible modification of the genome,” said the principal investigator of one of the studies, Amy Wagers, PhD, a professor of stem cell and regenerative biology at Harvard University.
Unlike currently devised exon-skipping systems using antisense oligonucleotides, which will likely require repeated dosing to remain effective, she said, CRISPR is a one-time process that does not require repeating.
All three studies showed that local delivery of the editing system using an adeno-associated virus (AAV) vector led to deletion of the mutant exon in up to 18 percent of muscle fibers, increased dystrophin expression in both skeletal microfibers and cardiac muscle, and improved muscle function including increased grip strength, without evidence of unintended changes in other genes. Systemic delivery caused a widespread increase in dystrophin production in myofibers, cardiomyocytes, and muscle stem cells.
While important issues of delivery, efficacy of editing, and safety remain to be worked out, “overall, the path to recovery using genome editing with CRISPR/Cas9 is poised for translation and looks very promising,” said Jerry R. Mendell, MD, director of the Center for Gene Therapy at Nationwide Children in Columbus, OH. Dr. Mendell, a leader of one of the ongoing exon-skipping trials in DMD, was not involved in any of the three studies in Science.
HOW CRISPR/CAS9 WORKS
CRISPR/Cas9 is a bacterial immune system, in which short nucleotide sequences from invading viruses are copied and stored in the bacterial chromosome in a region characterized by “clustered regularly interspaced short palindromic repeats” (CRISPR). These virally-derived sequences serve as recognition templates for future infections. When a sequence from a new virus matches an RNA copy of a stored CRISPR sequence, a nuclease called CRISPR-associated protein 9 (Cas9) cuts the viral DNA, inactivating the virus. Thus, the RNA serves as a sequence-specific guide, telling Cas9 where to cut.
The system used for gene editing, as described in the three Science DMD studies, provides a synthetic guide RNA, designed to match a target site on the mouse DNA. Genes for the guide RNA and Cas9 are delivered by AAV vector and taken up by target cells, where they are expressed. The guide pairs with the enzyme, and together they bind to the target sequence in the mouse chromosome, causing Cas9 to cut the chromosome.
“Gene editing at a fundamental level is creating breaks in the human genome, and then letting the cell repair it,” said a senior author of one of the studies, Charles A. Gersbach, PhD, an associate professor of biomedical engineering at Duke University.
In the DMD mouse model, the gene mutation is in exon 23, and all three groups targeted the mutation in similar ways. “The overall goal is to eliminate a DNA sequence that directs the splicing machinery to include the mutant exon, so the spliceosome will go on to the next exon” to create the final messenger RNA, said Eric Olson, PhD, a professor and chair of molecular biology at the University of Texas Southwestern in Dallas, who was an author of the third study.
(This same idea drives the exon-skipping strategies currently being explored in DMD.)
While a single break in a splice site would likely suffice, in the current studies, the three groups used two guide RNAs, introducing two breaks — one with the mutant exon, and the other at a splice site at one end of the exon. DNA repair enzymes then stitch the cut ends back together, eliminating part of the exon and the splice site that directs its inclusion in the mRNA. The result is a transcript that makes a shorter, but still functional dystrophin, akin to the dystrophin made in the Becker muscular dystrophy, which has milder symptoms.
RESULTS OF THREE STUDIES
All three research groups found that treatment led to expression of the modified dystrophin protein in muscle fibers, with the exact amounts differing depending on the delivery method and the post-treatment time point.
Dr. Olson saw an increase in fibers expressing dystrophin from 7 percent at three weeks to 25 percent at six weeks following local injection into the tibialis muscle, with total protein expression equally 53 percent of normal. Dr. Gersbach also injected the tibialis, and found that 67 percent of myofibers expressed dystrophin at eight weeks, at about 8 percent of the normal level.
“Because of significant differences in protocols, these figures cannot be directly compared to evaluate relative efficacy,” Dr. Gersbach said.
Dr. Wagers also looked at gene correction in satellite cells, the stem cells that divide to replace lost fiber nuclei. “If you edit the gene in satellite cells, you've created a pool of progenitor cells that will add new nuclei that carry the edited gene,” she pointed out. This effect has the potential to increase the degree of functional recovery over time. Her results showed that satellite cells were indeed affected by the treatment, and the edited cells went on to contribute to muscle regeneration.
WHAT LIES AHEAD?
Multiple issues remain to be resolved before the CRISPR/Cas9 system is ready for human trials in DMD. All three groups looked for signs of off-target editing, with no evidence seen at the top ten likely sequences elsewhere in the genome. But this was in mice, Dr. Wagers cautioned. “One wouldn't want to over-interpret these initial safety results.”
Nonetheless, Dr. Olson pointed out, “In all the studies done so far in mice, not just in DMD, to my knowledge no one has reported tumorigenesis or other pathogenic effects due to CRISPR-mediated mutagenesis, so I think this probably won't be a major issue going forward.”
Another concern is the unknown immunogenicity of the Cas9 protein, which, if it emerges as a concern, would likely require immune suppression at the time of delivery. Altogether, a CRISPR-based editing strategy might be applicable to 83 percent of DMD patients, but not the remaining group, whose disease is due to large deletions in the gene.
“I think this is an exciting proof of principle,” Jeffrey Chamberlain, PhD, a professor of neurology at the University of Washington in Seattle told Neurology Today. “The question that arises now is whether the efficiency will be increased to the point where it could be physiologically relevant.”
Dr. Mendell said other questions remain to be answered: For example, how much dystrophin production is enough, and can gene editing provide that level of correction? Regarding the first question, Dr. Mendell said, “There have been estimates that you would need anywhere from 15 percent up to 50 percent of normal [dystrophin production], but the truth is, no one really knows what the amount needs to be. As a clinician and as a gene therapist, I would target for a minimum of 20 percent of the normal level.”
“This is really early work,” Dr. Gersbach said, “and there is a lot of work to be done to optimize the system. That's the next thing we are going to be working on.” Strategies for improving efficiency may include using muscle-specific promoters for the delivered genes and choosing alternative targets to increase the efficiency of Cas9.
“It's important to remember this is pretty far off from a clinical application,” Dr. Chamberlain cautioned, “but it will be exciting to see the technology mature.”
Dr. Mendell was more optimistic. Assuming the issues of immunogenicity and off-target editing can be put to rest, he said, “I think we can get to a clinical trial within two years.”