Research Signal – Source date 2026-03-26
A cell that stops dividing can still affect the tissue around it. Under persistent stress, some cells enter a state called senescence: they no longer progress through the cell cycle, and many release a mix of signals that can alter nearby cells. In the brain, those signals may contribute to inflammation and tissue dysfunction. That possibility has raised an appealing question: could removing senescent cells slow neurodegenerative disease?
A March 2026 review in Biomedicines examines evidence on senescence in Alzheimer’s disease, Parkinson’s disease and multiple sclerosis, drawing on research published between January 2005 and June 2025. It brings together observations from human tissue, laboratory and animal studies, and early clinical work. Its central challenge is also the one any treatment would face: researchers need to identify which cells are senescent, what those cells are doing, and whether changing them helps people.
A stressed cell can change its neighbors
Senescence is a response to persistent cellular stress, including DNA damage and oxidative stress. In cells that normally divide, it involves a lasting halt to cell division. Senescent cells can also develop a senescence-associated secretory phenotype, or SASP: a collection of released substances such as inflammatory signals, growth factors and enzymes that affect surrounding tissue.
The SASP is not automatically harmful in every setting. Its effects vary with the cell type and the stress that triggered it; some secreted factors can support tissue repair, while prolonged signaling may disrupt tissue function. In the central nervous system, the review discusses how sustained signals could intensify inflammation, affect blood vessels and the blood–brain barrier, and interfere with the work of nearby neural cells. Because one cell can release signals that affect its surroundings, even a limited population may have effects beyond itself.
The term needs care when applied to neurons. The standard definition of senescence includes cell-cycle arrest, a feature suited to cells that can divide. Neurons are post-mitotic: they do not normally divide. Researchers therefore describe senescence-like features in neurons, rather than assuming that every senescence marker in a neuron means it has undergone the same process as a dividing cell.
The implicated cells differ by disease
In Alzheimer’s disease, the review highlights senescence-related features in astrocytes and microglia. These supporting and immune cells help maintain the brain environment and handle material such as beta-amyloid. Human postmortem tissue analyses described in the review found increased senescence-related signals in Alzheimer’s brains and an association between beta-amyloid accumulation and microglial signatures linked to reduced phagocytosis—the process by which cells take up and clear material. This is an association; it does not show that senescence began the disease. Experimental work suggests that altered microglial activity could make amyloid clearance less effective and that senescent-cell signals may destabilize synapses.
Parkinson’s disease presents a different set of vulnerable cells. The review describes senescence-related markers and secreted factors in human brain tissue, including in astrocytes in the substantia nigra, a region important to movement. Cell and animal experiments suggest that stress related to alpha-synuclein, toxins and other conditions can induce senescence-like changes in astrocytes, microglia and dopamine-producing neurons. In these models, the proposed consequences include increased oxidative stress and worsening damage to dopamine-related cells. The experimental mechanisms add context to the human observations, but do not establish the same sequence of events in patients.
In multiple sclerosis, the concern is partly about repair. Oligodendrocyte precursor cells, or OPCs, can mature into cells that help restore myelin, the insulating layer around nerve fibers. The review discusses experimental findings that senescence-related changes in OPCs may limit their maturation. In laboratory systems, factors released by senescent cells, including HMGB1 and CCL11, have inhibited this process. If repair capacity is reduced, myelin damage may be harder to resolve; the cited mechanisms are experimental evidence, not proof that senescence drives disability in people with multiple sclerosis.
These different roles matter for treatment. Removing a senescent astrocyte, a dopamine-producing neuron or an OPC would not have the same biological consequences. A broad intervention could also affect cells whose functions are useful. The review’s disease comparisons point toward a need to understand the identity and role of the target population before attempting to eliminate or alter it.
Identifying senescent cells is still difficult
No single marker reliably answers whether a brain cell is senescent. Markers such as p16, p21 and SA-β-gal can be useful clues, but they are not specific enough on their own. Inflammatory molecules such as IL-6 and IL-8 may be released by senescent cells, yet they can also appear when microglia or astrocytes are activated for other reasons. A positive inflammatory signal could therefore reflect ordinary neuroinflammation rather than senescence.
The review argues for combining evidence within the same cell: multiple molecular markers alongside features such as altered cell shape, cell-cycle status where relevant, and SASP components. Researchers can assess SASP-related gene expression with qRT-PCR and measure selected proteins with ELISA; transcriptomic approaches can examine wider patterns of gene activity. Single-cell RNA sequencing can distinguish cell populations that bulk tissue measurements blend together and may reveal cell states missed by simpler tests.
Computational approaches, including machine-learning tools and cell-aging clocks, may help researchers find patterns in these data. They remain research methods, however. A pattern detected in a dataset is not yet a validated clinical test that can reliably identify the relevant senescent cells in a living patient. Brain cell types differ, and the markers may vary by disease, region and stage.
Removal experiments strengthen—but do not settle—the causal case
Senescence could contribute to neurodegeneration, arise because tissue is already degenerating, or do both in a reinforcing cycle. Protein accumulation, oxidative stress and chronic inflammation can all create cellular stress. Senescent cells, in turn, may release factors that add to tissue dysfunction. The human observations summarized by the review cannot determine which came first or prove that senescence initiates disease.
Intervention experiments offer a stronger test than observing markers alone. In mouse models of amyloid and tau disease, removing cells marked as senescent reduced pathology and inflammation and improved some cognitive measures. Other preclinical studies reported lower amyloid burden or better aspects of synaptic function after senolytic treatment. These results support the possibility that senescent cells can influence disease progression in experimental systems. They do not demonstrate that removing such cells slows neurodegeneration in people.
There is also a practical risk in removing cells without enough selectivity. Senolytics aim to kill senescent cells; senomorphics aim to change their harmful activity without eliminating them. Either strategy would need to reach the relevant brain cells while avoiding damage to cells that are functioning or helping with repair. The review notes safety concerns, including non-selective neurotoxicity reported for navitoclax in human neuron models, alongside the broader difficulty of targeting specific cell populations.
A small Alzheimer’s pilot shows what early evidence can—and cannot—say
The review describes a pilot, open-label study of dasatinib plus quercetin in five people with mild Alzheimer’s disease. The treatment was tolerated over the 12-week study period, dasatinib was detected in cerebrospinal fluid, and investigators observed changes in inflammatory biomarkers and selected senescence mediators. Cognitive function did not change significantly over those 12 weeks.
This small study offers early information about whether the drugs reach the central nervous system and whether selected biomarkers change. It does not establish cognitive benefit or general treatment safety. With no reliable clinical test for the target cells, biomarker changes alone cannot show that treatment selectively removed senescent brain cells. Larger, carefully designed studies would need to show both that an intervention acts on the intended population and that this leads to meaningful benefit for patients.
What targeting the right brain cells could change
If researchers can reliably identify the harmful senescent cell populations and safely change their effects, treatments tailored to the cells implicated in each disease could help slow tissue damage or preserve repair capacity. That could matter to people living with neurodegenerative disease, but the path to that benefit depends on better cell identification, evidence of therapeutic effect in people and acceptable safety. The review describes a promising research direction, not a treatment whose patient benefit has been established.
Key Sources
- Primary source: Cellular Senescence in Neurodegeneration: From Cell Types to Therapeutic Opportunities.
DOI or Journal Page · Review or Meta-Analysis · 2026-03-26
Keep exploring