Base-Editing Approach Eases Huntington’s Disease in Mice by Silencing the Toxic Fragment

Researchers at the University of Illinois Urbana-Champaign have used a precise form of gene editing to ease Huntington’s disease in mice — not by switching off the culprit gene, but by changing how it is read.

Huntington’s is caused by an expanded stretch of CAG repeats in the HTT gene, which leads cells to produce a toxic, misfolded form of the huntingtin protein that damages neurons.

Why silencing the gene is problematic

Rather than silencing HTT entirely — which also does useful work — the team, led by Pablo Perez-Pinera and Thomas Gaj in the Department of Bioengineering, took a subtler route.

The huntingtin protein has genuine functions. It participates in transport of cargo along neurons, in the production of a growth factor supporting neuronal survival, and in development — complete loss is lethal in embryos.

How much reduction healthy neurons tolerate over decades is not established, and it is the central uncertainty of the silencing approach. Several programmes reducing huntingtin production have advanced clinically, and the question of whether long-term partial loss is safe remains open.

How base editing differs

The researchers designed base editors — tools that rewrite a single DNA letter without cutting the double helix.

Conventional CRISPR editing cuts both DNA strands and relies on the cell’s repair machinery to produce the intended change. That repair is error-prone, can produce unintended insertions or deletions, and occasionally causes larger rearrangements.

A base editor pairs a disabled cutting enzyme with a chemical modifier that converts one DNA base into another directly, guided to the right position but never severing the backbone. The change is precise and the repair pathway is not invoked.

For neurons the advantage is particularly relevant, since they do not divide and therefore cannot use the accurate template-based repair pathway that dividing cells rely on.

What they changed

The editors target exon 13 of HTT to disrupt a splice acceptor site. That prompts cells to skip the section encoding sites where the protein is cut into its most damaging fragments.

Splicing is how cells assemble a final message from a gene: sections are removed and the remainder joined, guided by sequences marking where each section begins and ends. Disrupting an acceptor site makes the machinery skip past that section entirely, producing a protein missing that portion.

The specific portion removed contains the sites where enzymes cleave huntingtin. The full-length protein is relatively tolerated; the fragments produced when it is cut are what accumulate and kill neurons. Removing the cleavage sites means the protein remains intact.

“Instead of turning the protein off completely, we alter how the gene is read so that the most damaging protein fragments are not produced,” Gaj said. To find the best design, the team screened more than 140 base editors.

What happened in mice

Delivered into the brain using adeno-associated viruses, treated mice accumulated fewer toxic protein fragments, showed fewer symptoms, and had less brain degeneration than untreated animals. Published July 29, 2026 in Nature Biomedical Engineering.

The three results form a chain: the molecular change occurred, the pathology it should prevent was reduced, and the animals were functionally better. Each step being present makes the mechanistic account considerably more convincing than any one alone.

The delivery constraint

AAV delivery to the brain is established and limited. The viruses do not cross the blood-brain barrier efficiently, so most approaches inject directly into brain tissue or cerebrospinal fluid — and coverage of a whole human brain, far larger than a mouse’s, is a genuine unsolved problem.

Huntington’s primarily affects the striatum, which is somewhat helpful since it defines a target region rather than requiring whole-brain distribution.

The caveats

This is early, preclinical research, and results in mice frequently fail to translate.

The authors say the next steps are humanised mouse models and larger-animal studies to assess tolerability and define a safe, effective dosing window before human testing could be considered.

Why Huntington’s attracts so many approaches

The disease has drawn disproportionate therapeutic attention relative to its prevalence, and the reasons are instructive.

The genetics are unambiguous. A single gene, a single type of mutation, complete penetrance — someone carrying an expanded repeat above the threshold will develop the disease if they live long enough. There is no uncertainty about the cause, which removes the target-validation problem that defeats most neurodegeneration programmes.

Patients can also be identified before symptoms begin. Genetic testing identifies carriers decades in advance, so a preventive trial is conceivable in a way it is not for conditions diagnosed only after damage has accumulated.

What has not followed is a treatment. Approaches have included antisense oligonucleotides reducing huntingtin production, RNA interference, and gene therapy, with clinical programmes that have variously been paused, discontinued or produced ambiguous results.

The gap between certainty about the cause and failure to treat it is what makes the disease a proving ground. A modality that works here would demonstrate something about the modality, because nothing about the biology is in doubt — which is why so many genetic-medicine platforms choose it as an early target.

Permanence is the property that cuts both ways. An edit is not reversible — if the modified protein proves to have problems over decades, there is no discontinuing it. Research news, not medical advice.