Cellular Senescence: When It Protects and When It Becomes a Problem

The phrase “zombie cell” is memorable, but it leaves out the most interesting part. A senescent cell is not half dead and is not always harmful. It has stopped dividing and changed jobs. Sometimes it protects us. Sometimes it helps repair. If it stays too long, it can become a difficult neighbour.

Cellular senescence is a stable response to signals such as DNA damage, dysfunctional telomeres, oncogenic stress or certain injuries. It belongs among the hallmarks of ageing, but it does not mean that one person is necessarily “ageing fast”.

The cell-cycle stop that protects

In 1961, Hayflick and Moorhead observed that normal human fibroblasts did not divide indefinitely in culture. Their original paper helped overturn the idea that every normal cell was immortal outside the body.

We now know that a cell can enter senescence through several routes. Very short telomeres are one, but DNA damage, oncogene activation, radiation, some treatments and tissue stress can also trigger it. Proteins such as p53, p21 and p16 may participate in cell-cycle arrest, with important differences by cell type and stimulus.

The outcome has a protective logic: a damaged cell stops copying itself. That barrier can reduce the chance that a dangerous alteration keeps spreading. Senescence also participates in embryonic development and specific phases of wound repair. Its existence is not the problem. Failure to resolve it after its job is done can be.

When persistence changes tissue

Senescent cells remain active. Many develop a senescence-associated secretory phenotype, known as the SASP. This is not a fixed recipe. It varies by tissue, trigger and time and may include cytokines, chemokines, growth factors and matrix-remodelling enzymes.

The work by Coppé and colleagues helped describe how these signals affect nearby cells. In the short term they can recruit the immune system or support repair. When persistent, they can maintain inflammation, disturb tissue structure and, in some contexts, support a harmful tumour environment.

This explains the apparent contradiction. The same family of responses can restrain a precancerous cell and later contribute to a damaging microenvironment if that cell is not cleared. Age adds two plausible problems: accumulated damage and less efficient immune surveillance.

How a senescent cell is identified

This is where enthusiasm for tests needs to cool. There is no exclusive label that switches on in every senescent cell and never appears elsewhere. The MICSE consensus initiative recommends combining evidence of stable arrest, structural changes, SA-beta-gal activity, p16 or p21 and SASP components, always within a tissue and experimental context.

Signal	What it may add	Why it is insufficient alone
p16INK4a	Involvement of a cell-arrest pathway	Not every p16-positive cell is senescent and not every senescent cell expresses p16
p21	A response linked to p53 and damage	It also appears in temporary arrest and other states
SA-beta-gal	A common lysosomal change in laboratory work	It depends on method and is not exclusive
SASP	Secreted signals and effects on surroundings	It is heterogeneous and overlaps with non-senescent inflammation
Lack of proliferation	Shows that the cell is not dividing	Quiescence and differentiation also stop division

Recent work describes different “senotypes” even within the same tissue. A 2025 human analysis reinforces this heterogeneity: populations defined by different markers do not overlap neatly. A supposed blood test of “whole-body senescent burden” cannot therefore be inferred from one molecule.

The controversy around antibodies that confuse p16^INK4a with another protein is a useful example. Our guide to p16 and the antibody problem explains why target, tissue, controls and independent validation need to be known before an image counts as evidence.

What animals show

The most striking functional evidence comes from mice. In a transgenic system, the study by Baker and colleagues removed p16-positive cells and observed improvements in several age-related problems, plus a longer median lifespan in that model.

This matters because it moves from association to intervention. Its limits are equally clear: the researchers used a genetic system built for the experiment, not an available pill; they removed cells defined by one signal; and the organism was a mouse. The study shows that certain cells can cause dysfunction in that model. It does not show that a healthy person should undergo a “senolytic cleanse”.

What we know in humans

Human evidence is much smaller. An open-label pilot gave dasatinib plus quercetin for three days to nine people with diabetic kidney disease. Eleven days later, several senescence-associated markers in adipose tissue and blood had fallen. This was a target-engagement signal, not a trial of rejuvenation, kidney function, prevention or longevity.

Negative results matter too. UBX0101, a local candidate for knee osteoarthritis, reached a phase 2 trial with 183 participants without a clear benefit on its primary pain endpoint. An attractive mechanism can be real while a particular drug, dose, tissue or disease still fails to respond.

The honest picture is that some early interventions affect human markers, but controlled trials have not yet established meaningful clinical benefit and an acceptable risk-benefit balance for healthy people.

Why self-treatment is not a sound translation

Dasatinib is a prescription medicine with important risks. Its combination with quercetin in a pilot does not make an over-the-counter quercetin product equivalent treatment. It also does not justify copying doses or schedules from a small study. Fisetin and other candidates follow the same principle: availability is not efficacy.

Our specific guide to senolytics and zombie cells reviews candidates, positive and negative trials and why no preventive schedule has been approved. Separating mechanism, activity on a marker and clinical benefit prevents a stimulating hypothesis from becoming premature advice.

What about everyday habits?

Exercise, adequate sleep, not smoking, good-quality food and cardiometabolic risk control improve health through well-established pathways. Some studies suggest that they also modulate senescence-related and inflammatory signals. That possible mediation is interesting, but we do not need to call them “natural senolytics” to recommend them.

The practical goal is not to chase a p16 score. It is to reduce harmful exposures, preserve function and treat known risks. If a senotherapeutic test or treatment eventually proves clinically useful, it will need to add value beyond those measures and clearly explain for whom, against which outcome and at what biological cost.

Frequently asked questions

What is a senescent cell?

It is a cell that maintains a stable halt in division and changes its function. It often responds to damage or stress, remains metabolically active and can send signals into its environment.

Does cellular senescence mean ageing?

They are not synonyms. Senescence is a cell programme that can occur at any age. Its accumulation and less efficient clearance do contribute to several processes associated with ageing.

Are all senescent cells harmful?

No. Senescence helps restrain damaged cells, participates in development and repair and can suppress tumours. Trouble begins when it persists, accumulates or disrupts tissue.

What is SASP?

It is the changing group of signals released by some senescent cells, including cytokines, growth factors and enzymes. It may recruit immune cells and support repair, but can also sustain inflammation or damage the surroundings.

Is there a clinical test for senescent-cell burden?

There is no validated routine test that measures senescent-cell burden throughout the body. p16, p21, SA-beta-gal and SASP signals are interpreted in panels and in a tissue-specific context.

Have senolytics been shown to rejuvenate healthy people?

No. Animal results are compelling and small human studies in disease show preliminary signals, but rejuvenation, prevention and longer life have not been demonstrated in healthy people.

Can I take quercetin or fisetin as a senolytic?

Availability as a supplement does not establish a safe or effective senolytic dose. Combination studies do not validate each component on its own or justify self-treatment.

Does exercise remove senescent cells?

Exercise improves many outcomes and may modulate inflammation and senescence-related pathways, but it should not be described as a clinically proven senolytic in humans.

Sources

1. Hayflick L, Moorhead PS. The serial cultivation of human diploid cell strains. 1961. PMID: 13905658.
2. López-Otín C et al. Hallmarks of aging: An expanding universe. 2023. PMID: 36599349.
3. Gorgoulis V et al. Cellular Senescence: Defining a Path Forward. 2019. PMID: 31788430.
4. Coppé JP et al. Senescence-associated secretory phenotypes. 2008. PMID: 19053174.
5. Baker DJ et al. Naturally occurring p16-positive cells shorten healthy lifespan. 2016. PMID: 26840489.
6. Hickson LJ et al. Senolytics decrease senescent cells in humans. 2019. PMID: 31542391.
7. ClinicalTrials.gov. Phase 2 study of UBX0101 in knee osteoarthritis. NCT04129944.
8. Ogrodnik M et al. Human senotypes and marker heterogeneity. 2025. PMID: 41162753.

This article is informational. No senolytic anti-ageing protocol is approved for healthy people, and medicines or supplements can cause adverse effects and interactions.