C9orf72 ALS: Mechanism, Neuron Pathology, and Disease Insights

A six-letter DNA sequence, GGGGCC, repeated hundreds or even thousands of times inside the C9orf72 gene is the most common known genetic cause of both amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). Healthy people carry fewer than about 30 of these repeats, but expansion carriers harbor anywhere from dozens to several thousand copies, and the mutation damages motor neurons through at least three converging mechanisms rather than one simple pathway. Understanding how a single repeat expansion can wreak so much havoc on the nervous system has reshaped how researchers think about neurodegenerative disease and opened new, if still imperfect, avenues for treatment.

How the C9orf72 Protein Normally Works

Before diving into what goes wrong, it helps to know what C9orf72 does in a healthy cell. The protein acts as a kind of traffic coordinator inside cells, physically interacting with a family of small signaling molecules called Rab proteins that direct cargo through the cell’s internal transport and recycling systems. Laboratory experiments have shown that C9orf72 associates with several of these Rab proteins and colocalizes with structures involved in autophagy, the process cells use to break down and recycle damaged components.¹ More specifically, C9orf72 helps regulate the very first step of autophagy by guiding one of the key initiation complexes to the right place in the cell. When C9orf72 levels are experimentally reduced in neurons, autophagy slows down and clumps of a waste-disposal marker protein called p62 accumulate, a pattern strikingly similar to what is seen in patient brain tissue.² In short, the normal C9orf72 protein helps keep the cell’s cleanup crew running. Lose half of it, and garbage starts piling up.

Three Roads to Neuron Damage

The field has settled on a framework in which the repeat expansion inflicts harm through three overlapping routes, often called loss of function, RNA toxicity, and dipeptide repeat protein toxicity. No single route seems sufficient to explain the full disease on its own; instead, they interact in ways that are still being untangled.

Loss of Normal C9orf72 Protein

The expanded repeat region can fold into unusual DNA structures called G-quadruplexes. These structures physically block the cellular machinery that reads the gene, reducing how much C9orf72 messenger RNA the cell produces.³ On top of that, the DNA around the expansion tends to become chemically modified through a process called methylation, further dialing down gene expression. Larger repeat expansions correlate with heavier methylation of the gene’s promoter region, and heavier methylation correlates with lower protein output.⁴ The result is that expansion carriers end up with roughly half the normal amount of C9orf72 protein, a state called haploinsufficiency. With autophagy and intracellular trafficking already compromised, neurons become less able to deal with the toxic products the same mutation is simultaneously generating.

Toxic RNA That Gets Stuck in the Nucleus

The expanded repeats are still transcribed into RNA, even if less efficiently. That repeat-laden RNA does not behave normally. Instead of being exported to the cytoplasm and put to work, it folds into tight tangles and accumulates as clumps called RNA foci inside the nucleus. Studies using patient-derived neurons have found that roughly a third of cells containing the expansion harbor these intranuclear foci.⁵ Making matters worse, the gene is read in both directions, producing sense and antisense transcripts, each of which forms its own set of foci in patient brain tissue.⁶

These RNA foci are not just inert blobs. They act like molecular flypaper, sequestering proteins that the cell needs for other jobs. One key protein trapped by the foci is hnRNP-H, a factor involved in processing other RNA messages. When RNA-binding proteins like hnRNP-H are soaked up by repeat RNA, normal gene regulation throughout the cell starts to go sideways.⁷ Antisense RNA foci appear to be particularly damaging in motor neurons: their presence in motor neurons tracks with mislocalization of TDP-43, the protein whose aggregation is considered the hallmark signature of ALS neurodegeneration.⁸

Dipeptide Repeat Proteins Made from “Untranslatable” RNA

Perhaps the most surprising discovery in this field was that the repeat RNA gets translated into protein even though it sits in a non-coding region of the gene. Through a noncanonical mechanism called repeat-associated non-AUG (RAN) translation, both the sense and antisense transcripts are read in all three possible reading frames, producing six different dipeptide repeat proteins (DPRs).⁶ Research has identified a specific start codon with a favorable surrounding sequence that initiates translation of at least one of these species, poly-GA, and knockout of a non-canonical translation factor called eIF2A impairs that production.⁹

Of the six DPR species, the two that contain arginine, poly-GR and poly-PR, are by far the most toxic. They carry dense positive charges that let them stick promiscuously to cellular components. But they behave differently from each other. Poly-PR tends to migrate into the nucleolus because proline spaces the arginine charges in a way that favors nuclear entry, whereas poly-GR, whose glycine backbone is more flexible, largely stays in the cytoplasm and binds tightly to a smaller number of molecular targets.¹⁰ These distinct behaviors mean each arginine-rich DPR damages the cell in its own way, even though both originate from the same repeat expansion.

How Neurons Break Down at the Cellular Level

The three toxic streams described above converge on several cellular systems at once, and the damage compounds over time.

Nuclear-Cytoplasmic Transport Goes Awry

A healthy cell maintains a strict border between the nucleus and the cytoplasm, shuttling molecules back and forth through pores in the nuclear envelope. The repeat expansion disrupts this gatekeeper system. In patient motor cortex tissue, a key pore-associated protein called RanGAP1 forms abnormal clumps instead of distributing smoothly around the nuclear rim, and these clumps colocalize with repeat RNA.¹¹ When trafficking between the nucleus and cytoplasm breaks down, proteins that should be in the nucleus end up outside it and vice versa, setting off a cascade of dysfunction in gene regulation and protein quality control.

Stress Granules That Refuse to Dissolve

Under stress, cells form temporary clusters of stalled RNA and proteins called stress granules. These are normally reversible. But arginine-rich DPRs undergo a process called liquid-liquid phase separation and trigger spontaneous stress granule assembly even without an external stressor.¹² Recent work has shown that these DPRs bind a critical stress granule scaffold protein called G3BP1 about a thousand times more strongly than RNA does, essentially hijacking granule assembly and driving a slow transition from liquid-like droplets into hardened, aggregate-like states that can recruit ALS-linked proteins such as TDP-43.¹³ This conversion from reversible droplet to irreversible aggregate may be one bridge between DPR toxicity and TDP-43 pathology, the two dominant features found together in patient tissue.

Mitochondrial Damage and Oxidative Stress

Poly-GR, the arginine-rich DPR that largely stays in the cytoplasm, has been found to preferentially bind to proteins of the mitochondrial ribosome, the machinery mitochondria use to make their own essential components. The result is compromised energy production, increased oxidative stress, and DNA damage inside motor neurons, with the oxidative burden growing worse with age.¹⁴ Motor neurons are especially energy-hungry cells, so mitochondrial dysfunction hits them particularly hard.

Nucleolar Stress Comes Early

One intriguing finding is that the nucleolus, the structure inside the nucleus responsible for building ribosomes, begins to shrink in motor neurons before TDP-43 has even mislocalized and before RNA foci are abundant. This suggests that nucleolar stress may be among the earliest events in the disease process, not a late consequence of it.¹⁵ Poly-PR, which preferentially homes to the nucleolus, may be partly responsible. In cell models, poly-PR self-assembly inside the nucleus has been shown to directly trigger condensation and aggregation of TDP-43, offering a mechanistic link between DPR accumulation and the hallmark proteinopathy of ALS.¹⁶

Why Motor Neurons Are Especially Vulnerable

ALS selectively ravages motor neurons despite the repeat expansion being present in every cell of the body. Part of the answer lies in excitotoxicity. Motor neurons from C9orf72 expansion carriers have been shown to express more of a specific glutamate receptor subunit that permits calcium to flood into the cell. This shift toward calcium-permeable receptors was found to be specific to motor neurons and absent in cortical neurons from the same patients.¹⁷ Excess calcium triggers a chain of events including enzyme activation and mitochondrial overload that can kill the cell outright.

Large-scale gene expression profiling has added another layer: neuron populations that are more resistant to the repeat expansion tend to have higher baseline activity of protein-homeostasis pathways, suggesting that neurons already well-equipped to handle misfolded or aggregated proteins survive longer, while those with thinner margins of quality control succumb first.¹⁸ Combined with the enormous energy demands and extreme physical length of motor neurons, these factors create a perfect storm of vulnerability.

The Role of Immune Cells in Neurodegeneration

Neuron death in C9orf72 ALS is not purely a cell-intrinsic problem. C9orf72 turns out to be expressed at its highest levels not in neurons but in myeloid immune cells, including the brain’s resident immune cells, microglia. When C9orf72 is lost or reduced in mice, microglia develop swollen waste-processing compartments and shift into a chronically activated, pro-inflammatory state that worsens with age, producing elevated levels of inflammatory signaling molecules.¹⁹ The inflammatory profile of these C9orf72-deficient mice resembles what is seen in C9orf72 ALS patient tissue more closely than it resembles sporadic ALS, pointing to a mutation-specific inflammatory environment.²⁰

Studies using microglia grown from patient-derived stem cells have confirmed this at a human level. Microglia carrying the C9orf72 expansion show impaired ability to engulf and clear debris, along with an exaggerated inflammatory response when stimulated, reinforcing the idea that immune dysfunction is a genuine contributor to neurodegeneration in these patients, not just a bystander effect.²¹

The Repeat Itself Keeps Growing

One under-appreciated aspect of C9orf72 disease is that the repeat expansion is not fixed. It continues to change size, both across generations and within the body of a single person over their lifetime. A mouse model engineered to carry 96 human G4C2 repeats revealed two distinct modes of expansion: small, incremental increases driven by the cell’s mismatch repair machinery, and large jumps triggered when DNA breaks occur near the repeat, relying on a different repair pathway. In living animals, the gradual mode was the primary driver of repeat growth in tissues over time.²²

This somatic instability has striking consequences. Autopsies of ALS and FTD patients have shown that repeat sizes vary enormously between tissues within the same person. In some cases, non-neural tissues carried fewer than 100 repeats while neural tissues harbored expansions 20 to 40 times larger.²³ This means that a blood test showing a modest expansion could be hiding a far larger one in the brain, and it raises the possibility that somatic expansion itself contributes to disease progression and the selective vulnerability of the nervous system.

Epigenetic Modification as a Double-Edged Sword

The methylation that silences the C9orf72 gene creates a paradox. On one hand, it reduces the amount of functional C9orf72 protein, which compromises autophagy. On the other hand, silencing the mutant gene also means less repeat RNA and fewer dipeptide repeat proteins are produced. Evidence suggests the second effect may outweigh the first, at least in some patients. In people with FTD carrying the expansion, higher promoter methylation was associated with later age at death and longer disease duration, and also with smaller repeat sizes.²⁴ This finding supports the idea that the toxic gain-of-function products, the RNA foci and DPRs, may be more damaging than the loss of C9orf72 protein itself, and that the body’s attempt to silence the mutant gene is at least partly protective.

Measuring Disease Before Symptoms Appear

One of the six DPR species, poly-GP, has emerged as a promising biomarker. Because it is produced in relatively large quantities and is chemically stable, it can be detected in cerebrospinal fluid. An ultrasensitive assay has achieved 100% specificity and 100% sensitivity for distinguishing C9orf72 expansion carriers from non-carriers, with the lowest signal from a carrier still running about eightfold higher than controls.²⁵ Critically, poly-GP has been detected in the spinal fluid of expansion carriers who have no symptoms yet, suggesting that DPR production begins years before disease onset and can serve as an early marker as well as a way to measure whether a therapy is hitting its target.²⁶

Therapeutic Approaches and Their Limits

Because the toxic RNA and DPRs flow directly from the repeat expansion, the most obvious strategy is to destroy the repeat-containing RNA before it can do harm. Antisense oligonucleotides (ASOs) are short synthetic DNA-like molecules designed to bind and trigger destruction of a specific RNA target. In preclinical mouse models, a single injection of a C9orf72-targeting ASO produced sustained reductions in RNA foci and DPR levels and improved behavioral deficits, establishing proof of concept for the approach.²⁷

Translating that success to humans, however, has been harder. BIIB078 (tadnersen), the first ASO targeting C9orf72 repeat-containing transcripts to reach clinical trials, was designed to degrade the toxic RNA via an enzyme-based cleavage mechanism while preserving the normal protein-coding messenger RNA.²⁸ In its Phase 1 trial, the drug did not reduce neurofilament levels, a marker of ongoing neuronal damage, and showed no clinical benefit relative to placebo, leading to discontinuation of its development.²⁹ Other ASOs and related strategies remain in development, and the failure of one specific molecule does not invalidate the underlying approach, but it underscored how far the field still has to go.

Alternative strategies are being explored in parallel. Small molecules that bind specifically to the G-quadruplex structures formed by the repeat RNA have shown the ability to reduce RNA foci and rescue cell death in both cell lines and fruit fly models.³⁰ At the most radical end of the spectrum, CRISPR-based gene editing has been used to delete the expanded repeat entirely from patient-derived stem cells, eliminating RNA foci and restoring normal methylation patterns without disrupting C9orf72 protein expression.³¹ Both of these are far from the clinic, but they demonstrate that multiple angles of attack are feasible in principle.

Spreading Between Cells

There is growing evidence that DPRs do not stay put in the cells that produce them. Using cell-culture platforms that included spinal motor neurons derived from C9orf72-ALS patients, researchers found that DPRs can spread from cell to cell through both vesicle-dependent and vesicle-independent pathways.³² If this happens in living patients, it could help explain why the disease progresses along anatomically connected regions of the nervous system over time, a pattern clinicians have long observed but struggled to explain mechanistically. Cell-to-cell DPR transmission would also mean that even neurons not carrying the expansion could be poisoned by their neighbors.

Modeling the Disease Outside the Human Body

Studying a disease that unfolds over years inside the human spinal cord presents obvious practical challenges. Patient-derived stem cell technologies have partially bridged this gap by allowing researchers to grow motor neurons and even more complex tissue structures from the cells of expansion carriers. Neuromuscular organoids, three-dimensional mini-tissues containing motor neurons, muscle fibers, and supporting cells, grown from C9orf72-ALS patient cells have been shown to develop muscle weakness, visible defects at the nerve-muscle junction, and loss of the insulating cells that normally wrap nerve fibers.³³ These organoid models reproduce multiple disease features in a dish, giving researchers a human-relevant platform to test drugs and probe disease mechanisms that would be inaccessible in animal models.