Instead of Switching the Huntington's Gene Off, Scientists Rewrote One Letter in Mice

Most Huntington's gene therapies aim to switch the toxic huntingtin gene off. A University of Illinois team tried something different: it used base editing, a CRISPR cousin that swaps a single DNA letter rather than cutting DNA, to reroute how mouse brain cells read the gene so the chunk of protein that gets clipped into toxic fragments is skipped entirely. The treated mice did better on every disease measure the team tracked.

The result, [published in Nature Biomedical Engineering](https://www.nature.com/articles/s41551-026-01747-y), is preclinical. It is also the first in vivo demonstration that base editing can be used to change how the huntingtin gene is read, rather than how much of it the cell makes, to treat a fatal inherited brain disease with no disease-modifying therapy.

Huntington's is a dominantly inherited neurodegenerative disease caused by a CAG trinucleotide repeat expansion in exon 1 of the huntingtin (HTT) gene. About 41,000 people in the United States have symptoms and more than 200,000 are at genetic risk, [per an estimate cited by Genetic Engineering & Biotechnology News](https://www.genengnews.com/topics/genome-editing/base-editing-strategy-alleviates-huntingtons-disease-in-mice/). The disease is fatal, and no therapy has yet been shown to slow its progression. That combination, fatal, heritable, and untreatable, is the reason a mouse paper still travels.

The dominant strategy in Huntington's drug development is to lower the amount of mutant HTT protein the cell produces, either by silencing the gene with RNA interference or antisense oligonucleotides, or by using CRISPR nucleases to disrupt it outright. Several programs built on that approach are already in clinical trials. The Illinois team, led by bioengineers Pablo Perez-Pinera and Thomas Gaj, took a different path.

Their idea was to leave the huntingtin gene mostly alone and instead edit how the cell processes its messenger RNA. Huntingtin protein is normally cleaved near its N-terminus, and in the mutant form that cleavage produces a small toxic fragment that accumulates inside neurons and drives disease. The team designed adenine base editors, enzymes that chemically change a single adenine to a guanine, to disrupt the splice acceptor at the start of HTT exon 13. When that splice site no longer works, the cell's splicing machinery skips exon 13 entirely, producing a shorter HTT protein that resists the toxic cleavage.

To find editors that did this with the fewest unintended edits, the team designed and screened more than 140 base editors, then packaged the lead candidates into adeno-associated virus (AAV) vectors and delivered them directly into the brains of mice carrying the mutant human HTT gene. Treated mice produced fewer toxic N-terminal fragments, showed fewer behavioral symptoms, and had less neurodegeneration than untreated controls.

"What we've shown is a new kind of potential treatment that demonstrates we can use genetic base editing to address a genetic disease by modifying how the protein is processed, not by inactivating the gene or by directly repairing the causal mutation," Gaj said in the University of Illinois News Bureau release. Perez-Pinera framed the work as a small edit that changes how the cell processes the protein, and stressed that without a cure in sight, the field needs "multiple shots on goal."

What the mouse study does not close is the list of questions that has dogged every Huntington's gene-therapy program. Lowering mutant HTT is necessary but has not been shown to be sufficient on its own, and the paper does not test whether the exon-13-skipped isoform retains normal huntingtin function over the long term. The team used AAV, the most common brain-delivery vehicle in the field, but AAV capsids do not reach every neuron in the human brain, and a single intracranial injection covers only a fraction of the striatum and cortex that degenerate in Huntington's. Base editing can also produce unintended edits elsewhere in the genome, and the team screened more than 140 candidate editors precisely to minimize that risk in this study. Whether the result holds up in humanized mice, in larger animals, and at the doses a human brain would require is the work the team has now set for itself.

The stated next steps are humanized mouse studies to test whether wild-type HTT stays above a tolerated threshold, large-animal tolerability and dose-range work to define a therapeutic window, and the development of non-AAV, less invasive brain delivery. A roughly 41,000-patient U.S. population waiting on a disease-modifying treatment will read the mouse numbers carefully, and the field will read them more carefully still: a mechanism that works in mice is news because the disease is fatal, and the next readouts will be humanized mouse and large-animal data, not a clinic date.