Five Kinds of Neurons Bear the Brunt of ALS Brain Damage

When Karin Danzer’s team at the German Center for Neurodegenerative Diseases began sorting through brain tissue from more than 80 deceased ALS patients, they were looking for patterns inside a disease that has, so far, refused to yield any. Not which neurons were dying (pathologists have known for decades that motor neurons go) but something more specific: which exact cell types, within a single brain region, were accumulating the toxic protein deposits that define the illness. The answer, published this week in Nature Communications, turns out to be cleaner than anyone quite expected. And considerably more complicated.

ALS is, in its essential shape, a disease of movement. It destroys the nerve cells that instruct muscles, working through the brain and spinal cord until breathing itself becomes impossible. In about 15 percent of cases it arrives alongside frontotemporal dementia, which strips away personality and language at the same time. There are no treatments that meaningfully change its course. Most patients survive between two and five years from diagnosis.

The molecular signature shared across almost all cases is a protein called TDP-43. Normally it lives in the nucleus, where it performs critical housekeeping duties on RNA, helping to suppress aberrant genetic sequences from being spliced into proteins that don’t belong. In ALS, TDP-43 abandons its post. It clumps in the cytoplasm of neurons, forming tiny aggregates visible under a microscope, and the RNA system begins to malfunction in its absence. Cryptic exons (genetic sequences that should be silenced) get incorporated into messenger RNA instead, producing garbled proteins and, apparently, contributing to cell death. This pattern appears across ALS patients regardless of the specific genetic mutation driving their disease, which is why the field has long regarded TDP-43 pathology as the convergence point, the place where many different roads arrive at the same bad destination.

What nobody had established, at least not in the motor cortex, was whether this convergence was uniform.

Danzer’s group worked with post-mortem tissue from patients in Germany, the Netherlands, Scotland and the United States: 30 with ALS, 20 with the mixed ALS-FTD form, 32 neurologically healthy controls. The motor cortex, the strip of cortical tissue responsible for voluntary movement, was their focus, partly because it’s the most clinically relevant area for understanding ALS-driven paralysis. “In ALS, as well as in the mixed form of ALS and FTD, TDP-43 deposits occur in different regions of the brain,” Danzer says. “However, the motor cortex is particularly relevant for movement disorders, which is why we focused on this area.” To get the kind of resolution they needed, the team combined three different molecular techniques: nuclear sorting by flow cytometry, single-nucleus multi-omic sequencing that reads both gene activity and chromatin accessibility simultaneously, and spatial transcriptomics that maps gene expression to physical locations within the tissue. It is, even by the standards of contemporary neuroscience, an unusual degree of methodological triangulation.

The result confirmed something suspected but never precisely demonstrated in this region. “The protein aggregates occur predominantly within excitatory cells,” says Danzer, “that is, within neurons that serve to transmit and amplify nerve signals. These cells seem to be particularly susceptible to the disease.” Selective vulnerability, the idea that particular neuron classes are disproportionately targeted even within the same tissue, has been an organizing concept in neurodegeneration research for years. It explains why Parkinson’s disease takes dopaminergic neurons in the substantia nigra, why Alzheimer’s strips entorhinal cortex first. In ALS, the principle had been invoked but rarely mapped at this resolution.

The more revealing finding, though, was the layer beneath. Within the excitatory neurons, five distinct subtypes emerged as the primary casualties four belonging to the class that sends projections within the telencephalon, the forebrain structures, and one from the extratelencephalic population, which projects its axons down into the brainstem and spinal cord. These long-range projecting neurons, sometimes called corticospinal-type, are precisely the ones you’d expect to be central to motor control and therefore perhaps central to ALS pathology. That they appear specifically vulnerable, rather than excitatory neurons en masse, suggests something about their molecular environment makes them uniquely susceptible; the transcriptomic data begins to hint at what.

Cryptic exon inclusion turned out to be cell-type-specific, not generic. Each of the five affected subtypes had its own distinct pattern of aberrant splicing, its own set of genes disrupted by TDP-43 dysfunction. This matters therapeutically, in ways that are somewhat uncomfortable to contemplate. If the disease were simply “TDP-43 goes wrong, all motor neurons die,” a single intervention restoring TDP-43 function might in principle suffice. The picture that emerges from this analysis is more demanding. “Our data offer insights into disease mechanisms and thus point to possible targets for therapy development,” says Danzer. “The observation that not all neurons are equally affected suggests that future therapies will need to be tailored to specific cell types in order to combat the disease effectively.”

There are real limits to what any post-mortem tissue study can tell you. The samples represent end-stage disease, or close to it, which means the researchers are reading a story at its final chapter rather than watching it unfold. Whether the five most-affected subtypes were vulnerable from the disease’s earliest stages, or became so through some cascade initiated elsewhere, cannot be determined from this dataset. The progression question how TDP-43 pathology propagates through cell populations over time remains genuinely open.

What the study does offer is a map. Not a complete one, but specific enough to be useful. The gene sets disrupted in each subtype could potentially serve as molecular targets points where an intervention might shore up a failing cell-type-specific process without needing to solve TDP-43 dysfunction globally. The spatial transcriptomic data, which anchors the molecular findings to actual locations in the cortical layers, adds another dimension: it shows not just which cells are affected, but where they sit in the architecture of the motor cortex, and what their neighbours are. “For example, one can see how the activity of certain genes is altered depending on the cell type,” Danzer notes. That granularity, rather than averaging gene expression across all neurons, is what makes the dataset potentially actionable.

ALS research has seen enough promising leads collapse on the way to the clinic that the field has become somewhat cautious about its own findings. The cellular specificity identified here will need to be validated in living disease models, and any therapeutic approach based on cell-type targeting will face the considerable additional challenge of reaching the right neurons with sufficient precision. But the precision of the question has changed. For decades the problem was posed roughly as: why do motor neurons die in ALS? The question becoming possible to ask now is more granular: why does this particular population of excitatory cortical neurons, sitting in layer five, projecting to the spinal cord, carrying this specific transcriptional signature, become vulnerable first? That is a different kind of problem. Harder to pose. Marginally easier, perhaps, to solve.

DOI / Source: https://doi.org/10.1038/s41467-026-69944-6

Frequently Asked Questions