The Friedreich Ataxia Trinucleotide Repeat Explained

Friedreich ataxia is caused by an abnormally long stretch of three DNA letters, GAA, repeated over and over inside a gene called FXN. Healthy people carry roughly 5 to 33 of these GAA repeats in the first intron of FXN, but people with the disease typically have hundreds, sometimes more than a thousand, stacked end to end on both copies of the gene. That bloated repeat does not scramble the protein’s blueprint directly. Instead, it jams the gene’s machinery so the cell can barely read it, starving the body of a protein called frataxin that mitochondria need to function.

Why a Repeat in an Intron Matters So Much

Introns are stretches of DNA that get snipped out before a gene’s message is translated into protein. They are sometimes called “junk,” which makes it seem strange that a repeat sitting inside one could cause a devastating disease. The answer is that the expanded GAA repeat does not need to end up in the finished protein to cause harm. It acts earlier, at the level of gene expression itself. When the GAA tract grows long enough, it triggers the cell’s chromatin-packaging system to lock down the surrounding DNA, converting the region from an open, readable state into tightly packed material that the gene-reading machinery can no longer access efficiently.

Researchers have found that cells from people with Friedreich ataxia show hallmark chemical tags on the proteins that package DNA near the FXN gene. These tags, including trimethylation of specific histone residues, recruit silencing proteins that further compact the region.

¹ In one study, cells from patients showed enrichment of heterochromatin-associated histone marks such as H3K9me3 and H3K27me3 at the start of the FXN gene, along with recruitment of the silencing protein HP1.² DNA methylation adds another layer of repression, and the combined effect is that FXN produces only a fraction of the frataxin a healthy cell would make.³

On top of the epigenetic silencing, the long GAA repeat can form unusual DNA structures in the test tube. When two long GAA tracts come together, they can create a self-associated structure sometimes called “sticky DNA,” which may physically impede the enzymes that transcribe the gene.⁴ So the expanded repeat attacks gene expression from multiple angles: structural roadblocks and an epigenetic shutdown that makes the DNA virtually inaccessible.

What Frataxin Does in the Cell

Frataxin is a small mitochondrial protein, and understanding what it does makes the disease’s symptoms easier to grasp. Mitochondria rely on iron-sulfur clusters for many of the enzymes in their energy-producing chain. Frataxin acts as an activator of the molecular complex that builds those clusters: it binds the assembly machinery and stimulates the enzyme that donates sulfur for the reaction.⁵ Structural studies have shown in detail how frataxin reshapes key loops within this assembly complex, unlocking an inhibitory zinc block and freeing up a critical cysteine residue so that sulfur transfer can proceed.⁶

When frataxin is scarce, iron-sulfur cluster production drops. Iron that was supposed to be incorporated into those clusters or into heme (the oxygen-carrying component of hemoglobin) instead accumulates inside the mitochondrion with nowhere to go.⁷ Excess free iron generates reactive oxygen species that damage the organelle’s membranes and DNA. The result is a vicious cycle: failing mitochondria, iron overload, and oxidative stress that slowly kills cells.⁸

Why the Heart and Nervous System Bear the Brunt

Friedreich ataxia is often described as a neuro-cardiodegenerative disease. The tissues hit hardest are those with the highest energy demands: the large sensory neurons of the dorsal root ganglia that relay information about limb position, the cerebellum and spinal cord tracts that coordinate movement, and the heart muscle. Dorsal root ganglion neurons appear especially vulnerable to frataxin deficiency, even compared to other cell types, with evidence of impaired mitochondrial function, reduced sirtuin activity, and abnormal protein acetylation in these cells.⁹ Cardiomyopathy is the leading cause of death in most patients. The shared thread is that these tissues depend heavily on mitochondrial energy production and are disproportionately harmed when that process fails.

Longer Repeats, Earlier Onset, Worse Disease

Not every person with Friedreich ataxia has the same number of GAA repeats, and the difference matters clinically. Each patient carries two expanded alleles (one inherited from each parent), and the lengths can differ. Studies have consistently found that longer repeats correlate with an earlier age of symptom onset, with the shorter of the two alleles being the better predictor. That shorter allele accounts for roughly half of the variation in when symptoms first appear.¹⁰ This makes biological sense: the shorter allele determines the maximum amount of frataxin the cell can produce, since it is less severely silenced than the longer one.

Repeat length also predicts specific complications. Patients with longer mean allele lengths are more likely to develop diabetes and cardiomyopathy.¹⁰ Still, the correlation is far from perfect. Two people with the same repeat lengths can have noticeably different disease courses, which tells us that other genetic and environmental factors are at play.

Interruptions in the Repeat Tract

One of those modifying factors is whether the long GAA repeat is pure or broken up by non-GAA sequences. Most expanded alleles are uninterrupted, which is part of what allows the unusual DNA structures and epigenetic silencing to take hold. But a minority of patients carry interruptions, short stretches where the GAA pattern is broken. When those interruptions sit near the end of the repeat tract (the 3′ end), they are associated with later disease onset. Modeling of this effect has estimated that a 3′ interruption delays symptom onset by about nine years compared to patients with pure, uninterrupted repeats of similar length.¹¹ These interruptions likely destabilize the abnormal DNA structures that would otherwise form, allowing more frataxin to be produced than the raw repeat count would predict.

Large interruptions, however, are very rare in the Friedreich ataxia population.¹² So while the concept helps explain some of the variability in disease severity, it applies to only a small subset of patients.

Compound Heterozygotes and the Rare Exceptions

About 96 percent of Friedreich ataxia patients are homozygous for the GAA expansion, meaning both copies of FXN carry a long repeat. The remaining fraction are compound heterozygotes: they have an expanded repeat on one allele and a different kind of mutation on the other, such as a point mutation, small deletion, or insertion. The clinical picture for these individuals depends on what the non-expansion mutation does to frataxin.

Compound heterozygotes whose second mutation completely eliminates frataxin production from that allele (“null” mutations) tend to have earlier onset and higher rates of diabetes compared to homozygous expansion patients. In contrast, compound heterozygotes whose second mutation allows some partially functional frataxin to be made show later onset and lower diabetes rates.¹³ This reinforces the general principle: the total amount of working frataxin in the cell is the central driver of how severe the disease becomes.

The Repeat Is Unstable Over Time

GAA repeats are not static. They can expand or contract during cell division, during transmission from parent to child, and even in cells that no longer divide, such as mature neurons.¹⁴ This somatic instability means that the repeat length measured from a blood sample may not perfectly represent what is happening in the brain or heart. It also means that the repeat can grow over a person’s lifetime in the very tissues that are most vulnerable, potentially contributing to disease progression.

Intergenerational instability is the reason repeats reach disease-causing lengths in the first place. A “premutation” or “mutable normal” allele in the range of roughly 34 to 65 repeats does not cause symptoms but is prone to expanding when passed to offspring. Over generations, these borderline alleles can jump into the pathological range. Once an allele exceeds a critical threshold, it tends to expand further in subsequent generations, a phenomenon sometimes described as genetic anticipation though the pattern in Friedreich ataxia is less dramatic than in some other repeat expansion diseases.

How the Repeat Is Tested

Diagnosing Friedreich ataxia genetically means measuring GAA repeat length on both alleles. Clinical labs typically use a tiered approach. A first-pass screening method, such as triplet-repeat primed PCR, can rapidly detect whether an expansion is present. This technique is highly sensitive and specific, with no false positives or negatives reported in validation studies, though it is limited in its ability to assign a precise repeat count to very large expansions.¹⁵

Samples that screen positive are then reflexed to a second method, such as long-range PCR on an agarose gel, for confirmation and sizing of the expanded alleles. One validated two-tier protocol classifies alleles into normal, mutable normal, borderline, and expanded categories, with the first-tier assay resolving alleles under about 200 repeats to within a few repeats, and the second tier sizing larger expansions to within roughly 50 repeats.¹⁶ This level of precision matters for genetic counseling and for stratifying patients in clinical trials.

Frataxin protein levels in blood have also emerged as a useful biomarker. A mass spectrometry-based assay measuring frataxin in whole blood found complete separation between healthy controls and Friedreich ataxia patients, with frataxin levels correlating with GAA repeat length and age of onset.¹⁷ This kind of biomarker is valuable not just for diagnosis but for monitoring whether a therapy is actually boosting frataxin production.

Friedreich Ataxia Among Repeat Expansion Diseases

More than 40 human diseases are caused by expanded DNA repeats, but Friedreich ataxia stands apart from most of them in two important ways. First, it is recessive: you need pathological expansions on both copies of the gene to develop the disease. Carriers with one expanded allele and one normal allele produce enough frataxin from their healthy copy and remain symptom-free. Most other trinucleotide repeat diseases, such as Huntington disease and myotonic dystrophy, are dominant, meaning a single mutant allele is enough to cause illness. Second, the Friedreich ataxia repeat causes disease through a loss-of-function mechanism: the expansion silences the gene so the cell cannot make enough of a needed protein. In contrast, many dominant repeat diseases involve a toxic gain of function, where the expanded repeat produces a harmful RNA or protein that actively damages the cell. Among repeat expansion disorders, loss-of-function mechanisms inherited recessively are uncommon.¹

This distinction has practical implications for therapy. Because the FXN gene itself is intact and the protein it encodes is normal, treatments do not need to neutralize a toxic product. They “just” need to coax the gene back on, or deliver working copies of frataxin, or compensate for the downstream metabolic damage. That is a fundamentally different therapeutic landscape from a disease where the mutant gene is itself the problem.

Therapies Aimed at the Repeat and Its Consequences

Because the expanded repeat silences FXN through epigenetic modifications, one therapeutic strategy is to reverse those modifications pharmacologically. A class of drugs called HDAC inhibitors (specifically, class I histone deacetylase inhibitors of the 2-aminobenzamide type) has shown the ability to reactivate FXN expression in patient-derived neurons, in mouse models, and in circulating immune cells of patients treated in an early-phase clinical trial.¹⁸ However, the response varies considerably from person to person. Research has linked this variability to the degree of DNA methylation at the FXN locus: heavily methylated gene copies appear resistant to reactivation, while partially methylated copies respond more readily.¹⁹

A different approach targets the expanded repeat directly with synthetic oligonucleotides. Antisense oligonucleotides complementary to the GAA repeat have been shown to activate frataxin protein expression in patient-derived cell lines at low nanomolar concentrations, and newer “gapmer” designs are several-fold more potent than earlier versions.²⁰ These compounds work by disrupting the abnormal RNA or DNA structures formed by the repeat, essentially prying the gene back open. They remain preclinical but the potency improvements are encouraging for eventual in-vivo testing.²¹

Gene therapy and gene editing represent more ambitious strategies. Researchers have explored delivering a working copy of FXN using viral vectors, and more recently have turned to CRISPR-based editing to either delete the expanded repeat or correct it in place. Animal and cell model studies have demonstrated proof of concept, though delivery to the right tissues in the right amounts remains a major hurdle.²²

The only approved drug so far takes a different angle entirely. Omaveloxolone (brand name Skyclarys), approved in the United States in February 2023, does not increase frataxin levels. Instead, it activates a transcription factor called Nrf2, which boosts the cell’s own antioxidant and mitochondrial-protection programs to compensate for the damage caused by frataxin deficiency.²³ In the pivotal trial, a dose of 160 mg per day improved neurological function scores relative to baseline, with the benefit growing over time compared to placebo.²⁴ A systematic review pooling data from two randomized trials confirmed time-dependent benefit, with the most noticeable improvements in upper-limb coordination and upright stability.²⁵ It is not a cure, but it is the first treatment to clear regulatory approval, and it validates the idea that downstream metabolic rescue can produce measurable clinical benefit even without fixing the repeat itself.

Challenges in Modeling the Disease

Developing treatments has been slowed by the difficulty of building good animal models. Mice do not naturally get Friedreich ataxia, and recreating the human disease in them has proved tricky. Completely knocking out mouse frataxin is lethal early in embryonic development. Models that reduce frataxin or introduce a human FXN gene carrying a long GAA repeat tend to produce either neurological or cardiac symptoms, but rarely both together in a realistic time course. Those that do develop both often have such severe, rapidly fatal disease that they do not reflect the gradual, decade-long progression seen in human patients.²⁶ This mismatch has made it hard to test therapies in a setting that predicts how well they will work in people.²⁷

More recent models have made progress toward capturing the slow, multi-system nature of the disease, and patient-derived cell lines, including induced pluripotent stem cells differentiated into neurons and heart cells, have become essential tools for testing whether a compound can boost frataxin in relevant human tissue. The combination of better animal models and human cell systems is gradually closing the gap between laboratory findings and clinical reality.