Amyotrophic lateral sclerosis (ALS) has long been viewed through the lens of motor neuron degeneration. Yet for nearly half of patients, the disease extends well beyond movement, producing impairments in executive function, language, behavior, and, in some cases, frontotemporal dementia. While ALS’s cognitive symptoms are increasingly recognized clinically, their biological basis has remained poorly understood.
A new multimodal study published in Cell suggests that cognitive decline in ALS is not driven by a single pathological process. Instead, different cognitive phenotypes appear to arise from distinct cellular programs involving specific combinations of neurons, glia, and vascular cells. The findings provide one of the clearest demonstrations yet that the cognitive manifestations of ALS are biologically heterogeneous—a realization with important implications for biomarker development and precision therapeutics.
Led by New York Genome Center and Columbia University Irving Medical Center researchers, the team combined spatial transcriptomics, single-nucleus RNA sequencing, chromatin accessibility profiling, and multiplexed imaging to construct a high-resolution atlas of two prefrontal cortex regions involved in cognition: the dorsolateral prefrontal cortex (BA46), which governs executive function, and Broca’s area (BA44/45), a critical center for language production. Importantly, all brain donors underwent standardized neuropsychological testing before death, allowing molecular changes to be directly associated with specific cognitive deficits.
The analyses revealed two fundamentally different biological signatures associated with cognitive decline.
Executive dysfunction localized primarily to deep-layer neuronal populations within the dorsolateral prefrontal cortex. These neurons exhibited reduced expression of genes involved in oxidative phosphorylation, mitochondrial function, synaptic organization, and neurotransmission, suggesting that impaired energy metabolism and disrupted neuronal communication are central features of executive decline. The affected neuronal populations also demonstrated altered mitophagy pathways and changes in glutamatergic signaling, consistent with widespread metabolic stress.
Language impairment was associated with a fundamentally different pattern of pathology. Rather than centering on selective neuronal dysfunction, language deficits correlated with a diffuse multicellular response involving reactive astrocytes, inflammatory microglia, endothelial cells, and vascular-associated transcriptional programs spanning multiple cortical regions. These findings suggest that language dysfunction may arise from altered interactions within the broader cellular microenvironment rather than dysfunction of a single neuronal population.
One of the study’s central conclusions is that cognitive decline reflects altered interactions among multiple cell types rather than dysfunction within a single vulnerable population. Across cognitively impaired brains, reactive astrocytes, activated microglia, oligodendrocyte precursor cells, neurons, and vascular cells shifted in coordinated cellular states, pointing to disease processes that emerge from disrupted multicellular networks rather than isolated cellular pathology.
The analyses also identified changes in the perivascular microenvironment that may have important implications for disease progression. Patients with cognitive impairment exhibited altered activation and spatial organization of microglia surrounding blood vessels, adding to growing evidence that dysfunction of the neurovascular unit contributes to neurodegeneration. These observations further expand the view of ALS beyond motor neurons alone, implicating interactions between immune cells and the cerebral vasculature in cognitive decline.
Notably, one of ALS’s defining pathological features—TDP-43 aggregation—did not fully account for the observed cognitive phenotypes. Although TDP-43 pathology was strongly associated with neuronal and synaptic transcriptional changes, it did not reliably predict which patients developed cognitive impairment. In contrast, many of the glial and vascular signatures associated with language dysfunction appeared largely independent of TDP-43, suggesting that multiple biological processes contribute to the clinical heterogeneity of ALS.
The study also illustrates the growing power of spatial biology to bridge clinical phenotypes with underlying molecular mechanisms. Rather than relying solely on bulk tissue analyses or isolated cell populations, integrating spatial transcriptomics with single-cell genomics enables researchers to identify not only which cells are altered, but how they interact within intact tissue architecture—a critical step toward understanding complex neurological disease.
Together, the findings reinforce a broader shift underway in neuroscience. Rather than viewing neurodegenerative diseases through the lens of single pathogenic proteins or vulnerable neuronal populations, increasingly detailed spatial and single-cell atlases are revealing them as disorders of interacting cellular networks. In ALS, that perspective suggests that different cognitive syndromes arise from distinct biological mechanisms—an insight that could ultimately guide more precise biomarkers, improved patient stratification, and therapies tailored to the biology driving an individual patient’s disease.
