ABOVE: In patients with muscular dystrophy, muscle fibers (purple) are replaced by adipose tissue (black) as muscles waste away and lose function. (Mesenchymal tissue shown in green and yellow.)  

If Chengzu Long hadn’t been quite so unlucky, he might never have attempted to study and treat Duchenne muscular dystrophy. As a PhD student in Eric Olson’s lab at the University of Texas Southwestern Medical Center, Long had spent years knocking out genes in mice to try to identify their role in muscle development and disease, only to find that each of the resulting knockouts had no discernible differences from wildtype individuals.

In the fall of 2013, with only about a year left until his planned graduation, Long decided to take a different approach: rather than generate yet another knockout mouse that might again lack a phenotype, he would use the new CRISPR-Cas9...

Like many muscle disorders, or myopathies, Duchenne is caused by mutations in a single gene—the X chromosome’s DMD gene, which encodes the protein dystrophin. Without a functional dystrophin protein, Duchenne patients gradually lose their mobility as their muscles degenerate. Most die by their early 30s from breathing complications or heart failure. “Duchenne is devastating,” says Olson, whose lab has worked on muscle development and disease for 30 years. “Right now there [are] 300,000 boys in the world with Duchenne, so it’s a large patient population, and there’s a desperation for a really transformative therapy.”

Because the genetics of the disease are well understood, researchers could theoretically replace the mutated version of DMD with a healthy copy to cure the disease. Unfortunately, the gene for dystrophin is massive, with 2.6 million base pairs. As a result, it’s not feasible to insert the entire gene, or even just the 11,000 coding base pairs (introns excluded), into a viral vector that could deliver the therapeutic package to the muscle. “Gene editing therefore was a great opportunity to correct the endogenous gene rather than trying to deliver” a nonmutated version of it, says Charles Gersbach, a biomedical engineer at Duke University.

By the end of 2013, multiple labs had successfully used gene editing to rescue the dystrophin protein in vitro, using cells from patients. So Long decided to try his luck at using CRISPR-Cas9 to edit the dystrophin gene in vivo. He injected the CRISPR system into the zygotes of mdx mice, which carry a single mutation in the gene for dystrophin. He then implanted the zygotes into female mice, and confirmed in their 10-day-old progeny that CRISPR-Cas9 had successfully corrected the mutation that causes Duchenne symptoms. When the corrected mice were a month old, Long tested their muscle function and found that it had also improved compared with mice carrying the uncorrected mutation. “We published one of the first in vivo rescues of phenotype [using CRISPR] in an animal model,” he says.

Now, just four years later, Olson has improved the gene-editing strategy to address the majority of human Duchenne mutations, and he and others have successfully tested these techniques in various mouse models, 3-D–engineered heart muscle, and dogs. With multiple research groups gearing up for clinical trials over the next few years, what they learn will have implications beyond Duchenne, as similar approaches could be applied to treat many other muscle disorders.

Dystrophin is part of a protein complex linking the cytoskeleton of muscle fibers to the surrounding connective tissue (basal lamina). It’s a long protein with numerous redundant coils (purple balls), and acts like a shock absorber during contraction. Without functional dystrophin to support muscle strength and stability, muscle fibers are easily damaged. See full infographic: WEB | PDF
© steve graepel

Mini proteins offer partial improvements

After successfully correcting the Dmd mutation in mdx mice during embryonic development, Long and Olson decided to use adeno-associated virus 9 (AAV9) to deliver the CRISPR system into mice after birth. They were again able to correct the gene mutation that led to the Duchenne phenotype, resulting in major improvements in muscle function.2 Gersbach at Duke and Amy Wagers, a stem cell biologist at Harvard University, simultaneously published similar results.3,4 “For the very first attempt, I personally thought the efficiencies were remarkably good,” says Wagers.

There are a few reasons why successful gene edits, once made, were so efficient in these examples. Muscle cells are multinucleated, with each cell having hundreds of nuclei. “If you can correct just a few of them you can protect the whole muscle fiber,” says Gersbach. And just a little dystrophin can go a long way toward improving muscle function, Olson adds. “It’s estimated that as little as 15 percent of normal dystrophin levels could be curative, or at least highly beneficial.”

What’s tricky, though, is restoring the defective parts of the gene in the first place. Doing so relies on a template-driven DNA repair process called homology-directed repair (HDR), which occurs infrequently in nondividing cells such as those of skeletal and heart muscle. “The problem with inserting stuff is that it’s very inefficient and makes the drug more complicated,” says Gersbach. “We’re exploring a number of ways by which we might increase that efficiency, but for the time being that’s not really an option.”

CRISPR-mediated exon skipping improves the function of heart muscle cells taken from Duchenne patients.

As an alternative, researchers can use CRISPR to initiate a different DNA repair process called nonhomologous end joining (NHEJ), which doesn’t rely on a template and is far more efficient than HDR. NHEJ can’t replace a defective gene with a wildtype version, but it can direct the removal of problematic portions of it. And there is good reason to think that such a strategy might work for Duchenne.

In Duchenne, DMD mutations disrupt the gene’s reading frame, causing translation to terminate prematurely and leading to a complete lack of a functional dystrophin protein. In a closely related disease called Becker muscular dystrophy (BMD), patients carry mutations in the DMD gene that are in-frame, typically deletions that result in a smaller but still partially functional dystrophin. As a result, patients with BMD generally suffer less-severe symptoms and survive considerably longer than Duchenne muscular dystrophy patients. “The dystrophin protein is built like a shock absorber with a series of redundant coils in the center,” says Olson. “You can delete several of those coils and still retain function.”

To improve Duchenne patients’ prognoses, then, researchers can provide them with smaller versions of the dystrophin protein. One promising approach is to scale down gene therapy so that the DNA encoding a functional, pared-down protein can fit into a viral vector. A recent trial testing the implantation of such a “microdystrophin” showed that the therapy increased levels of the small protein in muscles and reduced levels of a Duchenne-associated enzyme, called creatine kinase, in three patients. The results are promising, although it’s still too early to know what their clinical significance will be. “There are a number of these types of trials that are ongoing that look really exciting,” says Gersbach.

Alternatively, researchers can omit mutated exons to avoid the premature termination of translation. A few biotech companies have been testing antisense oligonucleotides to enact an “exon-skipping” strategy, obviating the need to edit the genome. The antisense oligonucleotides bind to the mRNA produced from mutated exons and cause them to be skipped during translation, restoring the transcript’s reading frame to produce a smaller but functional protein. The US Food and Drug Administration (FDA) recently approved a drug, Sarepta Therapeutics’s eteplirsen (Exondys 51), that uses antisense oligos to skip exon 51.

But the drug would only work on 13 percent of Duchenne patients—those with mutations in that exon—and so far the clinical benefits appear relatively modest. The drug is also expected to cost about $300,000 a year. In addition, the effects of the antisense oligos are transient, and patients will need regular injections to maintain the exon skipping. “There’s a real need for a long-term therapy that can go at the cause of this disease, which is the genetic mutation,” Olson says.

That’s where CRISPR-based genome editing may be able to help. Independent studies by Olson, Gersbach, and Wagers initially used a double-cut strategy—cutting on each side of an exon to excise it from the genome and glue the cut ends back together using NHEJ. But Olson recently developed an approach that uses a single cut to skip, rather than excise, a defective exon, bringing the protein back in frame. He used CRISPR-initiated NHEJ to alter a genomic region before the defective exon, causing the exon to be skipped during later splicing of the mRNA. “We spent a lot of time trying to conceive of the simplest possible approach to modify the genome to correct it, and that led us to single-cut CRISPR,” he says.

I think it’s not unreasonable to imagine that we could get into patients in a few years.

—Eric Olson, University of Texas Southwestern Medical Center

Olson and Long used the single-cut strategy to skip exon 51 in a mouse model of Duchenne and restore up to 90 percent of dystrophin protein expression in skeletal muscles and the heart.5 But again, translating this to humans would only affect about 13 percent of boys with Duchenne. To test a strategy that could rescue dystrophin function in a majority of patients, the researchers simultaneously targeted the top 12 exons that are “hotspots” for DMD mutations. By making a single cut before each of these exons, the team restored dystrophin protein expression in heart muscle cells, or cardiomyocytes, derived from patient stem cells.6 “We found, quite amazingly, that with CRISPR-edited cardiomyocytes the force of [heart muscle] contractions really recovered after the genome editing,” says Long.

Skipping multiple exons in the dystrophin gene—in particular, those with the highest rates of Duchenne-causing mutations—will be key to developing a broadly applicable therapy, says Olson, who says he believes his team is getting close. “There are more than 4,000 mutations that have been identified in Duchenne patients worldwide, so one of the huge challenges has been to devise a strategy that would allow you to consolidate large numbers of patients with different mutations and correct them with a common method,” he says. “We believe we can, in principle, correct somewhere between 60 and 80 percent of Duchenne mutations using this single-cut gene-editing strategy.”


Duchenne muscular dystrophy results from mutations in the DMD gene that encodes the dystrophin protein. There are many types of mutations that can cause the disease; each disrupts the reading frame such that translation terminates prematurely, producing no functional dystrophin protein. DMD mutations are particularly common in “hotspot” areas of the gene (exons 45–55 and 2–10).

Because the genetic underpinnings of Duchenne are known, researchers can devise gene-editing fixes to the problem. Several potential treatments are now being tested in preclinical and clinical studies. In some cases, they aim to correct the mutations in the DMD gene; other times, the goal is simply to restore a shorter but still partially functional dystrophin protein.
© steve graepel

From pipette to patient

If all goes well, researchers will soon test Duchenne gene editing in nonhuman primates, and then in human patients. Gersbach is collaborating with Sarepta Therapeutics to develop CRISPR gene editing for boys with the condition, while Olson’s work has led to the creation of Exonics Therapeutics, a biotech focused on designing gene editing therapies for Duchenne and other neuromuscular diseases. Last week, Olson and colleagues published a study in which they used AAV9 to deliver CRISPR to the muscles or blood of beagles with a naturally occurring DMD-like mutation. After treatment, biopsies of the dogs’ skeletal and heart muscles revealed increased dystrophin expression and improved integrity when viewed under a microscope.

But the biggest challenge is likely to be getting the therapy into patients, says Long. Treating Duchenne requires correcting both skeletal muscle—to restore mobility—and cardiac muscle, to prevent heart failure. This will require systemic delivery of any therapy through the blood. That’s worked so far in mice, but it remains to be seen how well it will work in humans. “Delivery into patients is a huge, huge challenge,” Long says.

The hope is that technologies devel­oped for Duchenne can be applied for other myopathies.

—Charles Gersbach, Duke University

Despite the long road ahead, these cutting-edge Duchenne therapies have advanced at a blistering pace over the past few years, and the future looks bright. “There’s no question that there will be challenges coming, but I’ve never been more optimistic about something,” says Olson. “I think it’s not unreasonable to imagine that we could get into patients in a few years.”

And Duchenne could be just the beginning. There are hundreds of incurable muscle and heart disorders that are caused by mutations in a single gene. “Duchenne is the obvious test bed,” says Gersbach, “but the hope is that technologies developed for Duchenne can be applied for other myopathies.”

Beyond Duchenne

In 2015, 
Ronald Cohn used CRISPR to remove duplicated exons, restoring the full-length dystrophin protein in cells from a Duchenne patient. Now, the chief pediatrician at the Hospital for Sick Children in Toronto is testing whether CRISPR can treat another myopathy, congenital muscular dystrophy type 1A (MDC1A).

MDC1A affects one in 150,000 people worldwide and is caused by a mutation in a gene called laminin alpha 2. It can lead to severe muscle wasting, paralysis, and death by the age of 30. Unlike the gene for dystrophin, laminin alpha 2 lacks redundant regions, so researchers can’t just excise or skip mutated exons or create a truncated but functional form of the protein.

In a mouse model of MDC1A, Cohn and his team used AAV9-delivered CRISPR to make two strategic cuts to repair a mutation in a splice site, thus restoring a skipped exon and the full-length protein. The treatment significantly improved the mice’s muscle strength and function (
Nat Med, 23:984–89, 2017). Cohn hopes to test this treatment in MDC1A patients in the next few years, and sees even more potential in the future.

“It’s in theory applicable to any disease with a splice-site mutation, and there are about 50,000 splice-site mutations associated with diseases,” he says. Cohn is currently testing the technique on tissues from Duchenne patients with rare splice site mutations, with promising results.

Other myopathies may also benefit from modern technologies. Molecular biologist
 Jocelyn Laporte at the Institute of Genetics and Molecular and Cellular Biology in France has been working for decades on rare and severe muscle disorders known as centronuclear myopathies, which can lead to death in the first year or two of life. Laporte and his team noticed increased levels of dynamin 2 in mice lacking the disease-associated gene Mtm1, and in muscle biopsies from centronuclear myopathy patients. The researchers used genetic crosses to produce Mtm1 knockout mice with only one Dnm2 allele (the double knockout is lethal), reducing dynamin 2 by 50 percent and successfully improving both the muscle function and longevity of diseased mice (J Clin Invest, 124:1350–63, 2014).

Injecting mice early in life with antisense oligonucleotides or small hairpin RNAs to reduce dynamin 2 expression prevented disease development (
Nat Commun, 8:15661, 2017). Injecting them later reversed some symptoms, including decreased muscle strength and mass (Mol Ther, 26:1082–92, 2018). “Reducing dynamin 2 really seems to have therapeutic potential,” says Belinda Cowling, a former postdoc with Laporte and currently head of research at biotech company Dynacure, launched in 2016 to translate the work to the clinic.

Laporte and Cowling hope to start clinical trials in 2019 using antisense oligonucleotides to reduce human dynamin 2 expression. “It’s quite an exciting moment to be working on centronuclear myopathies,” says Cowling. “And given its role in membrane remodeling, you could imagine that dynamin 2 may play a role in other diseases, so targeting it could also have broader therapeutic potential.”

Sandeep Ravindran is a freelance science writer living in New York City.


  1. C. Long et al., “Prevention of muscular dystrophy in mice by CRISPR/Cas9–mediated editing of germline DNA,” Science, 345:1184–88, 2014.
  2. C. Long et al., “Postnatal genome editing partially restores dystrophin expression in a mouse model of muscular dystrophy,” Science, 351:400–403, 2016.
  3. C.E. Nelson et al., “In vivo genome editing improves muscle function in a mouse model of Duchenne muscular dystrophy,” Science, 351:403–407, 2016.
  4. M. Tabebordbar et al., “In vivo gene editing in dystrophic mouse muscle and muscle stem cells,” Science, 351:407–11, 2016.
  5. L. Amoasii et al., “Single-cut genome editing restores dystrophin expression in a new mouse model of muscular dystrophy,” Sci Transl Med, 9:eaan8081, 2017.
  6. C. Long et al., “Correction of diverse muscular dystrophy mutations in human engineered heart muscle by single-site genome editing,” Sci Adv, 4:eaap9004, 2018.

Update: Since the print publication of this article, Eric Olson of the University of Texas Southwestern Medical Center and colleagues published a study testing their CRISPR-based treatment in dogs. A description of this study has been added to the online version of the story. 

Interested in reading more?

Magaizne Cover

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?