When David Liu first heard about a strain of mouse from his colleague Zheng-Yi Chen, he got excited. The mice carry a gene, TMC1, with a mutation that leads to deafness over time, giving them the name Beethoven mice. Their mutation matches one in humans that produces the same effect. The mutation is dominant; if it is present, hearing loss is certain.

Liu, a chemical biologist at the Broad Institute, works with the noted CRISPR-Cas9 technology, which targets and changes precise stretches of DNA. In the Beethoven mouse, he saw an ideal testing ground for the new gene editing technology, bringing hope that it might accomplish something new: improve hearing by disrupting a single genetic mutation. Other forms of gene modification add copies of genes, but are ineffectual if a dominant mutation remains.

In a paper published Dec. 20 in Nature, Liu, who is also a professor at Harvard University and an investigator of the Howard Hughes Medical Institute, and Chen, a hearing biologist at Massachusetts Eye and Ear Infirmary and professor of otolaryngology at Harvard, along with colleagues, report that the idea worked. The mice treated showed improved hair cell survival and hearing thresholds, and were startled by loud noises while untreated mice weren’t. “To our knowledge, this is the first time that genome editing has been used to correct hearing loss in an animal model of human genetic deafness,” Liu says. “There is a lot of work to do to translate these results into patients, but there is some proof of principle here.”

Until the advent of gene editing, geneticists could catalogue mutations without being able to do much about them. “David Liu’s work is a wonderful development in a growing trend that after all these years of trying we have been able to do something clinically meaningful about disease-causing or disease-predisposing changes in the genome,” says Fyodor Urnov, associate director of the Altius Institute in Seattle, Washington and a pioneer in gene editing who wrote a commentary that accompanies the study, but wasn’t involved in the research. “This work is exciting because it provides the first but essential step in advancing an approach like this to the clinic for this genetic condition.”

Hearing requires the conversion of acoustic energy into electrical signals. Sound waves travel through the ear and wash over the hair cells of the inner ear, which bend under pressure and send an electrical impulse up the auditory nerve to the brain.

Nearly half of all cases of deafness are due in some part to genetic factors, and many of those gene mutations affect the functioning of hair cells. The most common cause of genetic hearing loss, accounting for 20 percent of cases, is a recessive connexin 26 mutation on the GBJ2 gene. A recessive disease mutation requires a copy of the mutation from both parents. By contrast, one parent can pass along a dominant disease mutation like the one in the TMC1 gene, cause of 4 to 8 percent of cases of genetic hearing loss. TMC1 creates a defect in a protein that helps convert sounds into electrical signals, while the healthy copy of the gene is simply ignored. For people with the condition, hearing loss begins in childhood and deafness ensues within 10 to 15 years.

Gene editing and gene therapy are not the same thing. The latter involves the insertion of a good copy of a gene to overcome a deficiency in the copy that a patient is born with. One such clinical trial has begun for hearing loss that aims to regenerate hair cells in patients with acquired hearing loss by injecting a new gene. “It kickstarts supporting cells into becoming hair cells again,” says Lawrence Lustig, chair of otolaryngology at Columbia University Medical Center, one of three sites involved in the trial. But this approach doesn’t work for dominant mutations like the TMC1 mutation, says Lustig. “You can’t just grow new hair cells because you have the same bad genetic background that will kill those hair cells again.” (The Food and Drug Administration approved on December 19 a gene therapy for a rare inherited form of blindness. The treatment may cost more than $1 million.)

A technology like CRISPR-Cas9 is different. “The gene editor goes into the nucleus, finds the gene of interest and then asks are you normal or are you mutant? If it’s normal, it largely leaves it alone. But if it’s mutant, it cuts it and knocks it out,” says Urnov.

Liu and his colleagues targeted the mutant TMC1 gene copies by first binding the Cas9 protein to RNA guide molecules that program Cas9 to find and disrupt the target gene. Then they injected those protein-RNA complexes into the ears of newborn Beethoven mice.  Exquisite precision was required because the two copies of the gene—alleles—are so similar. “You can’t get closer than differing by a single base pair,” says Liu. To avoid disrupting healthy alleles, the team used an innovative method of delivery into the cell. Instead of the usual virus or DNA that programs the targeted cell to generate Cas9 and the guide RNA, they used a cationic lipid to directly deliver the Cas9 protein and the guide RNA. Liu likened the lipid to “fancy, tiny soap bubbles.” This method allows the editing agents to naturally disappear once their work is done, minimizing unintended damage to normal copies of the TMC1 gene. The process was roughly 10 times more precise than the viral delivery method, so that 20 copies of the faulty gene—rather than two—were disrupted for every normal copy that was effected.  

Although Liu speculates that the number of cells that were ultimately altered was a modest fraction of those in the inner ear, the effect was surprisingly strong. The thresholds at which the mice could detect a sound improved from 75 or 80 decibels (the noise level of a garbage disposal, say) to 60 decibels (normal conversation). The scientists aren’t sure why this “halo effect” protected surrounding cells, but it is an encouraging finding for the next step of experimenting with larger animal models such as nonhuman primates, whose anatomy is more like ours. A little bit of gene editing, it seems, can go a long way.

Because CRISPR-Cas9 can be guided to any gene, rewriting DNA with gene editing is akin to rewriting software. Other forms of deafness attributable to an errant copy of a single gene might also be ameliorated using the same technique. Altogether such cases amount to about 20 percent of genetic deafness.

Lustig, who works with patients with hearing loss every day, says Liu’s results are “significant” and offer hope for gene editing as a treatment. “It’s not around the corner,” he says, “but we’re on the pathway.”