Tools like CRISPR, which cut DNA to change its sequence, are moving incredibly close to the clinic as a treatment for some genetic diseases. But far from the spotlight, researchers are increasingly excited about an alternative that leaves the DNA sequence unchanged. These molecular tools target the epigenome, the chemical labels that adorn DNA, and the surrounding proteins that control gene expression and how it ultimately behaves.
Numerous studies in the last few years in mice have suggested that epigenome editing is a potentially safer and more flexible way to turn genes on or off from DNA editing. In an example described last month at a gene therapy meeting in Washington, DC, an Italian team reduced gene expression in mice to lower cholesterol levels in animals for months. Other groups are studying epigenome editing to treat everything from cancer to pain to Huntington’s disease, a fatal brain disorder.
Unlike DNA editing, where changes are permanent and may involve unintended results, epigenomic edits may be less likely to cause harmful effects outside the target and may be reversed. They can also be finer, slightly increasing or decreasing the activity of the gene, instead of blowing it up with full force or deleting it completely. “The exciting thing is that there are so many different things you can do with technology,” said Charles Gersbach, a longtime epigenome editing researcher at Duke University.
The addition or removal of chemical labels on DNA and histone proteins that coil around (see illustration, p. 1035) can either silence a gene or expose its sequence from DNA bases to other proteins that include it. Some cancer drugs remove or add these chemical labels, but as disease fighters they have had limited success. One of the problems is that drugs are out of focus, acting on many genes at once, not just those related to cancer, which means they come with toxic side effects.
But editing the epigenome can be done precisely, using the same enzymes that cells use to turn their genes on and off. Researchers are attaching key components of these proteins to a gene editing protein, such as a “dead” version of CRISPR’s Cas9 protein that can target a specific location in the genome but can’t cut DNA. Their effects can vary: one editor can remove histone markers to include a gene, while another can add methyl groups to DNA to suppress it.
Two decades ago, the biotechnology company Sangamo Therapeutics designed an epigenome editor using this method, which discovered a gene called VEGF, which helps promote the growth of blood vessels, in the hope of restoring blood flow in people with diabetes neuropathy. The company injected DNA encoding the editor into the leg muscles of about 70 patients in a clinical trial, but the treatment did not work very well. “We couldn’t deliver it effectively to muscle tissue,” said Fyodor Urnov, a former Sangamo scientist now at the Institute of Innovative Genomics at the University of California, UC, Berkeley.
So the company turned to adeno-associated virus (AAV), a harmless virus long used in gene therapy to efficiently deliver DNA to cells. The idea was that the cell’s protein-making machine would use DNA encoding an epigenome editor to ensure a stable supply of it. This strategy seems more encouraging: For the past three years, Sangamo has reported that mice can reduce levels of tau in the brain, a protein involved in Alzheimer’s disease, and levels of the protein that causes Huntington’s disease.
Other teams working with mice have used the AAV delivery approach to increase abnormally low protein levels to treat hereditary obesity, as well as Dravet’s syndrome, a severe form of epilepsy. Last year, a group used epigenome editing to rule out a gene involved in the perception of pain for months, a potential alternative to opioid drugs. Another team recently included a gene with an epigenome editor delivered by a virus other than AAV. They injected it into young rats exposed to alcohol; alcohol suppressed gene activity, which in turn left animals anxious and willing to drink. The epigenome editor reawakened the gene and relieved symptoms, the team said in May Scientific achievements.
AAVs tested by many groups are expensive and these DNA carriers, along with the foreign proteins they encode, can elicit an immune response. Another disadvantage is that the DNA strand encoding the epigenome editor is gradually lost in the cells as they divide.
Last month, at the annual meeting of the American Society for Gene and Cell Therapy in Washington, DC, genetic editing experts suggested an alternative to avoiding the shortcomings of AAV. A key step for the group led by Angelo Lombardo at the San Rafaele Teleton Institute of Gene Therapy came in 2016, when he, Luigi Naldini and others reported in cell that adding a cocktail of three different epigenome editors to the cells in a petri dish suppresses gene expression and that this continues until the cells divide.
This meant that instead of relying on AAVs to transfer DNA to their epigenome editors – and forcing endless expression – they could use lipid nanoparticles, a type of fat bubble, to carry their plan as messenger RNA (mRNA). In this way, the cells produce the protein for only a short time, which is less likely to elicit an immune response or make changes to the epigenome in unforeseen places. Such nanoparticles are considered safe, especially after being injected into hundreds of millions of people in the last two years to supply mRNA for COVID-19 vaccines.
It took a few more years for the Italian team to turn their laboratory research into an animal success. At the genomics meeting, postdoc Martino Capelutti of the Lombardo Laboratory described in detail how the team injected mice with fat particles carrying mRNA encoding epigenomic editors designed to silence a living gene. PCSK9which affects cholesterol levels. The strategy works by injecting 50% of PCSK9 protein into the bloodstream and reducing low-density lipoprotein or “bad” cholesterol for at least 180 days.
“I see it as a huge step forward,” said Urnov, who hopes the lipid nanoparticle approach will soon be extended to other genes in the disease. “The key thing here is that you don’t have to have a constant expression from the epigenome editor,” said Jonathan Weissman of the Whitehead Institute. Weissman – led work reported last year in cell on improved CRISPR-based epigenomic editors that make lasting changes.
Researchers say that editing the epigenome can be especially helpful in controlling more than one gene, which is more difficult to do safely with DNA editing. It can treat diseases such as Dravet’s syndrome, in which a person produces some of the necessary protein, but it is not enough, because as a dimer of light, the strategy can modulate gene expression without turning it on or off completely. Several new companies are hoping to commercialize the treatment with the help of epigenome editors. (Gersbach and Urnow founded one, Tune Therapeutics; Lombardo, Naldini, and Weissman were among the founders of another, Chroma Medicine.)
Despite the excitement, researchers warn that it will take time to edit the epigenome to have a wide impact. Editors don’t always work as advertised on some genes, says UC Davis epigenetics researcher David Segal. This may be due in part because, as epigenetics researcher John Stamatoyanopoulos of the University of Washington, Seattle, worries, researchers do not understand exactly what editors do after entering cells. “It’s a black box,” he says.
However, Stamatoyanopoulos agrees that editing the epigenome has a “huge promise.” Researchers now need to refine their epigenome editors, test them on other disease genes and tissues, and test them in larger animals for safety before moving them to humans.