A molecular-editing tool that’s small enough to be delivered to the brain shuts down the production of proteins that cause prion diseases, a rare but deadly group of neurodegenerative disorders.

The system — known as coupled histone tail for autoinhibition release of methyltransferase (CHARM) — changes the ‘epigenome’, a collection of chemical tags that are attached to DNA and which affect gene activity. In mice, CHARM silenced the gene that produces the disease-causing proteins in most neurons across the brain without altering the gene sequence.

This system is the first step towards developing a safe and effective ‘one and done’ treatment for reducing the levels of harmful proteins that cause prion disease, says Madelynn Whittaker, a bioengineer at the University of Pennsylvania in Philadelphia. The findings were published today in Science¹.

‘Epigenetic’ editing cuts cholesterol in mice

“The system addresses significant challenges faced with previous epigenetic-editing systems,” says Whittaker, who co-authored an accompanying perspective article in Science. Challenges include reducing the toxicity of editing tools and delivering them to cells without compromising their potency, she adds.

Prion diseases are caused by misfolded prion proteins (PrPs), which clump together and destroy neurons. This can lead to conditions such as fatal familial insomnia — a rare genetic disease that prevents people from sleeping and results in death. Although prion diseases are incurable, drugs known as antisense oligonucleotides (ASOs) have shown some promise. These short, single-stranded molecules bind to faulty messenger-RNA sequences and boost or reduce protein expression. Previous studies in mice infected with misfolded versions of PrP have shown that ASOs reduce expression of these proteins and extend lifespan². But the drugs require several injections to produce a long-term therapeutic effect and can result in adverse effects, such as liver damage, says Whittaker.

In 2021, Jonathan Weissman, a biochemist at the Massachusetts Institute of Technology in Cambridge, and his colleagues developed CRISPRoff — an editing tool that adds a chemical tag, called a methyl group, to the DNA strand, which reduces gene activity without altering the genome. But the tool cannot be delivered to brain cells, because its genetic components are too large to fit into an adeno-associated virus (AAV) — a common vehicle for ferrying gene therapies inside cells. “The real challenge was delivery,” says Weissman.

To address this, Weissman and his colleagues developed CHARM, which uses molecules called zinc-finger proteins to guide itself to target genes. These proteins are small enough to be delivered in an AAV vector.

The researchers tweaked CHARM to recruit and activate components of DNA methyltransferases — molecules found inside cells that add methyl groups to DNA, which alters gene expression. This reduces the toxic effects associated with adding molecules that originate outside the cell, says Weissman. “The only thing we changed in the cell was its ability to express the prion protein,” he says.

When the researchers delivered CHARM to the brains of healthy mice, they found that it reduced PrP expression by more than 80% across the entire brain — much more than the minimum level required to produce a therapeutic effect. Weissman and his team also engineered CHARM to switch itself off after it had finished its gene-silencing work, which prevented it from making copies of itself that could lead to harmful off-target effects.

The team behind CHARM includes Sonia Vallabh, a prion scientist at the Broad Institute of MIT and Harvard in Cambridge, who inherited the mutation behind fatal familial insomnia, and her husband Eric Vallabh Minikel. Twelve years ago, Vallabh and Minikel switched careers to investigate treatments for fatal familial insomnia. Vallabh says CHARM brings her “tremendous optimism”. She adds that drug development is typically slow, but the work demonstrates how quickly new approaches can be developed with the right team. “The amount you can accomplish in a short time is incredible,” says Vallabh. “It was only two years and one month ago that we first approached Jonathan with the notion of working together, and here we are.”

CHARM also has the potential to treat other diseases that are caused by the build-up of abnormal proteins, such as Parkinson’s and Alzheimer’s, adds Weissman. “We know that epigenetic silencing works for not every gene, but the majority of genes,” he says.

Jacob Goell, a researcher who develops epigenome-editing tools at Rice University in Houston, Texas, is optimistic that CHARM will one day land in the clinic. But more extensive work is needed to assess how the tool and the changes it creates interacts with cells’ genetic machinery, especially over longer periods, he adds.

The next step is to investigate how CHARM will work in an AAV vector that can target neurons in the human brain. “That’s the next big challenge,” he says.