Unlike germline variants, which are inherited in every cell, somatic variants arise post-zygotically and are unevenly distributed across brain regions, cell types and even individual neurons. This patchwork enables focal biological effects that can scale up to network-level dysfunction. Because each variant is private to a lineage or even a single cell, its influence depends critically on when during development it arose. Early embryonic events can generate broad regional clones and severe malformations, whereas variants arising later are confined to smaller populations of cells, producing more restricted lesions.

The developmental logic is now well illustrated in epilepsy. Somatic mutations activating the PI3K-AKT3-mTOR pathway cause hemimegalencephaly and focal cortical dysplasia, while mutations in genes such as SLC35A2 and components of the Ras-MAPK pathway underlie drug-resistant focal epilepsies. Recent cell-type-resolved studies have shown that clonal dynamics and cell-type-specific selective pressures shape mosaic architecture in the human forebrain, with lineage-tracing experiments revealing how neocortical clones distribute across lobes and hemispheres. Developmental timing, in other words, writes an anatomical map into the genome itself.

Ageing adds a second layer. Neurons, which do not divide and must survive for decades, accumulate DNA damage from oxidative stress and the energetic demands of signalling, and studies of single human neurons show mutation burdens rising steadily with age. Contrasting patterns have been documented between ageing neurons and oligodendrocytes, and some analyses point to age-associated hypermutability in a subset of brains. These private variants might cumulatively erode cellular resilience, weakening neurons’ capacity to withstand further stress and lowering the threshold for disease.

Neurodegenerative diseases show distinct increases in neuronal somatic variant burden, and the review highlights convergent mutational signatures implicating oxidative stress and topoisomerase 1, or TOP1, mediated DNA repair as shared mechanisms of genomic instability. Work on single Alzheimer’s disease neurons has documented diverse somatic genomic changes, and related alterations have been reported in chronic traumatic encephalopathy and in a range of tauopathies, suggesting that distinct disorders may leave characteristic genomic fingerprints in the cells they damage.

A separate axis of mosaicism involves microglia, the brain’s resident immune cells. Clonal haematopoiesis of indeterminate potential, or CHIP, describes age-related expansion of blood cell clones carrying mutations in genes such as TET2 and DNMT3A, and evidence now indicates that CHIP-associated clones can populate the brain as microglia. These mutant microglia adopt inflammatory and sometimes proliferative states and have been linked to Alzheimer’s disease and other neurodegenerative conditions, although mouse studies show the effects can be context-dependent, with some TET2-mutant myeloid cells appearing protective. Clonal inflammatory microglia have also been implicated in histiocytosis-associated neurodegeneration, underscoring that mosaicism in glia may be as consequential as mosaicism in neurons.

Perhaps the most provocative proposal in the review is that somatic variants confined to disease-relevant brain regions could act as focal initiating lesions. In diseases such as amyotrophic lateral sclerosis and frontotemporal dementia, recent sequencing studies have identified focal mutations associated with widespread degeneration, and somatic TARDBP variants have been reported as a cause of semantic dementia. The authors suggest that a local genetic insult could trigger pathology, including protein aggregation, that then propagates across anatomically connected brain regions, explaining the characteristic focal onsets and asymmetric spreads seen in Parkinson’s disease and ALS.

Detecting these variants is technically demanding, and the review takes stock of the rapidly evolving toolkit. Bulk sequencing with matched controls, error-corrected duplex sequencing and single-cell whole-genome amplification methods each trade sensitivity for coverage, and machine-learning callers such as DeepMosaic and DeepSomatic are improving detection across platforms. Single-cell approaches now extend to long-read sequencing and to single-cell transcriptomes, enabling de novo variant calling and the resolution of structural variants, retrotransposon activity and repeat expansions, including the somatic CAG repeat expansions that drive Huntington’s disease.

Clinical translation remains early. Mosaicism complicates genetic diagnosis because a variant detectable in blood may be absent from the affected brain tissue, and vice versa, and cerebrospinal fluid liquid biopsy is emerging as one route to sampling brain-specific mosaicism non-invasively. Yet the framework already has therapeutic echoes: mTOR inhibitors such as everolimus and sirolimus, developed for tuberous sclerosis, are being tested against focal cortical dysplasias driven by somatic mTOR mutations, demonstrating that mosaic lesions can be drug targets.

Zhou and Walsh argue that the next phase requires integrating genomic, cellular and physiological analyses in longitudinal human studies to establish causality, identify biomarkers and guide targeted interventions. If somatic mosaicism proves to be a genuine driver of neurodegeneration rather than a by-stander, medicine may need to treat the ageing brain not only as an organ under metabolic and inflammatory attack, but as a shifting genetic mosaic whose private mutations accumulate over a lifetime, quietly reshaping the vulnerability of each neuron and glial cell one DNA lesion at a time.

Subject of Research: Somatic mosaicism in the human brain across development, ageing and neurodegenerative disease

Article Title: Somatic mosaicism in the brain: linking development, ageing and neurodegeneration

Article References: Somatic mosaicism in the brain: linking development, ageing and neurodegeneration. (n.d.). https://doi.org/10.1038/s41582-026-01268-x

Image Credits: AI Generated

DOI: 10.1038/s41582-026-01268-x

Keywords: somatic mosaicism, brain, neurodegeneration, ageing, epilepsy, Alzheimer's disease, clonal hematopoiesis, microglia, DNA damage, single-cell sequencing, neurodevelopment, TOP1