DNA methylation at DUX4 linked to milder FSHD in nine-generation family
Higher levels of DNA methylation at the DUX4 region are associated with milder symptoms in facioscapulohumeral muscular dystrophy (FSHD), according to a study of a family spanning nine generations. The work, published in Brain, also identifies a DNA change near a little-studied gene, HDGFL3, that appears to be linked to higher methylation. The findings may help explain why people who carry the same disease-causing change can have very different outcomes.
A puzzle of variable severity
FSHD progressively weakens muscles of the face, shoulders and limbs. Its cause has been understood for years: a specific genetic change allows DUX4, a gene normally active only in early embryonic development, to switch on in adult muscle. This triggers toxic processes that damage muscle tissue.
Carrying the change means developing the disease, but severity varies widely. The article describes two siblings with the same genetic change. The brother noticed weakness in his early teens and could barely walk by his 50s, while the sister was symptom-free into her late 30s and only mildly affected throughout her life. Methylation, a chemical modification that can reduce gene activity or silence a gene altogether, had long been suspected as an explanation, but had not been clearly demonstrated.
Reading a difficult stretch of DNA
The team, led by Russell Butterfield (professor of pediatrics and neurology) and Robert Weiss (professor of human genetics) at University of Utah Health, drew on a family that has worked with the university's geneticists since the 1940s. Stored DNA samples date back to the 1980s, and these could be matched with historical clinical records.
The region containing DUX4 is hard to analyse because it is packed with long, repetitive sequences. Butterfield described it as the “gnarliest gene out there”. The researchers therefore built a targeted long-read sequencing tool able to resolve the region and measure methylation at single-base resolution. Comparing those measurements with clinical outcomes showed that more methylation at DUX4 went with milder disease.
A second gene enters the picture
The researchers then compared each participant’s methylation level with their whole-genome sequence, searching for shared DNA patterns linked to methylation differences. People with higher methylation at DUX4 were more likely to carry a specific DNA change near HDGFL3, a gene that no one had previously connected with FSHD.
The function of HDGFL3 is not well understood. Its structure suggests it may interact with DNA and help regulate methylation. The study therefore presents it as a candidate modifier of severity and a target for further work, not as a proven mechanism.
Implications for trials and therapy
The authors see several uses for the results:
- Prognosis: methylation at DUX4 could help predict how the disease will progress, supporting long-term planning for patients.
- Clinical trials: a clearer estimate of untreated progression could make it easier to detect whether a therapy works.
- Drug targets: treatments might raise HDGFL3 activity or increase methylation at DUX4 directly, suppressing the disease-causing gene.
Butterfield said a clinical trial aimed at increasing methylation at DUX4 is under way, and that its first gene therapy patient was treated a few weeks before the report. No results from that trial were given.
Limitations and next steps
The study is built on a single, extensively documented family, which is a strength for tracing inheritance but means the findings need testing in other FSHD cohorts. The HDGFL3 association is a statistical link, and the gene’s role in regulating methylation has yet to be shown experimentally. Weiss acknowledged that the problem is not solved. Priorities include replicating the methylation–severity relationship in wider populations and testing whether HDGFL3 influences DUX4 methylation in laboratory models.
Source: Genetic 'dimmer switch' can protect against severe disease in rare muscular dystrophy (Medical Xpress, 17 November 2026).
Image: AI-generated illustration.
Working on this topic? Find a journal for your research or join the discussion on Scholars Post.


