Research Signal – Source date 2024-11-22
Senescent cells are a tempting target for ageing research. They can accumulate in tissues and contribute to disease, so removing them might seem like a direct route to better health. But senescence is not simply a label for cells that have become useless.
A review published on November 22, 2024, brings together recent work on how these cells are identified, studied in animals, measured in people and targeted by therapies. Its central lesson is practical: useful treatment will depend on knowing which cells are involved, where they are, what they are doing and whether an intervention changes a meaningful health outcome.
Why removing every senescent cell would be the wrong goal
Senescence can contribute to chronic disease, including fibrosis, cancer and neurodegenerative disease. The review also describes roles for senescent cells in normal development and tissue repair. That dual role makes the therapeutic question more difficult than simply asking how to eliminate them.
A neonatal mouse lung example makes the point. In mice developing under normal oxygen conditions, a population of senescent cells appeared early after birth and declined between postnatal days 0 and 7. These cells showed neither an increase in the DNA-damage marker γH2AX nor an increase in the inflammatory markers examined. Mice exposed to excessive oxygen instead developed senescent cells with higher levels of inflammatory markers such as Il-1α, Il-1β, Cxcl2, Cxcl12 and Tnfrsf1b.
The researchers concluded that the early population in normally developing lungs was important for lung development. The comparison does not show that a senolytic drug would harm adult human tissue repair. It does show why a universal clearance strategy would be biologically crude: cells sharing a broad senescence label may have different triggers, locations, timing and roles.
That is the first boundary around the field’s promise. The goal is not to remove every cell that looks senescent. It is to distinguish disease-contributing senescence from forms that support normal biology.
A marker is a clue, not a complete identity
The next problem is deciding which cells belong in either group. A gene signature is a pattern of genes whose activity changes together. Researchers can compare that pattern with reference samples to assess whether a cell population resembles a particular form of senescence. The signature is evidence about cellular state, not a complete description of what the cell is doing in a tissue.
The review uses EndoSen to show why cell type matters. Researchers developed the signature from human endothelial colony-forming cells exposed to replicative, drug-induced or radiation-induced senescence. It combined 79 upregulated genes with 209 downregulated genes. When applied to endothelial cells, EndoSen showed a stronger enrichment score than broader signatures such as SenMayo and CellAge. That comparative advantage did not carry over to fibroblasts.
The result is useful precisely because it is limited. A pattern developed for cells lining blood vessels may be more informative in those cells than in another tissue. Treating it as a universal senescence test would discard the biological difference the experiment revealed.
The review therefore emphasizes combinations of markers rather than reliance on one feature. Transgenic models built around a single marker can be useful for tracking or manipulating a population, but they cannot capture every form of senescence. More complete identification may require several markers interpreted together with the tissue, the trigger and the surrounding disease context. Even then, a gene-signature match does not by itself establish that the cells are causing a patient’s disease or that they are ready to be treated.
Animal interventions test different therapeutic ideas
Senotherapeutics generally fall into two broad categories. Senolytics are compounds intended to induce the death of senescent cells. Senomorphics alter the senescence-associated secretory phenotype, meaning the signals and substances released by those cells, without necessarily removing the cells themselves.
Genetic mouse models have provided some of the strongest evidence that removing selected senescent populations can improve particular outcomes. In one approach described by the review, a tamoxifen-inducible system linked to the p21 gene allowed researchers to detect p21-positive cells, isolate them and, in a related model, genetically ablate them. Removing those cells improved physical function in ageing and obesity models. Another p21-based model distinguished p21-high cells from p16-high cells and reported that monthly clearance of p21-high cells in mice from 20 months of age extended lifespan and function in several tissues, including the heart and liver.
These experiments are important, but they test a precisely engineered biological intervention. They do not automatically predict what a drug will do in a person. A pharmacological treatment must reach the relevant cells, act in the intended tissue and avoid unwanted effects elsewhere. It may also affect a broader set of cells than the genetic system selected.
Fisetin illustrates the problem. The review reports beneficial effects in progeroid and naturally ageing mouse models, as well as in experimental muscular dystrophy models. Yet fisetin did not extend lifespan in genetically heterogeneous UM-HET3 mice in the Interventions Testing Program. The difference does not erase the positive findings, but it prevents them from being treated as a general property of the drug across models and endpoints.
Genetic cell removal and drug treatment therefore provide related but different tests of the same therapeutic idea. The first can show what happens when a defined population is manipulated. The second must establish whether a compound can reproduce that effect safely and reliably in a more variable biological setting.
The hardest measurement problem is knowing what the blood is telling you
A treatment cannot be evaluated well if senescence cannot be measured at the relevant site. A biopsy samples one location at one moment. It may miss other areas of the same organ or fail to represent a disease that is distributed unevenly. It also cannot, by itself, provide a complete picture of the body’s senescence burden.
Blood-based measurements are easier to repeat, but they introduce a different problem. Proteins associated with the senescence-associated secretory phenotype are not specific to senescence. Similar molecules can reflect ageing, cancer, Alzheimer’s disease, cardiovascular disease or other biological processes. A circulating concentration may therefore be associated with a patient’s outcome without showing that disease-causing senescent cells at a particular site produced it.
This difference matters when judging a treatment. A marker that helps predict morbidity is not necessarily a marker that demonstrates the removal of clinically relevant senescent cells. To make that connection stronger, researchers need to compare peripheral signals with local tissue measurements and examine how both change when senescence is experimentally reduced.
The review describes efforts such as the NIH Cellular Senescence Network, which is mapping senescent cells with single-cell technologies across tissues and lifespans in mice and humans. It also points toward more accessible samples, including saliva and urine, while stressing that a reliable combination of markers and a dependable way to track senolysis in living people are still needed. Better measurement is the infrastructure that allows a treatment result to be interpreted; it is not itself proof of clinical benefit.
A human trial narrows the promise
The review’s discussion of a phase 2 randomized trial shows how much precision is needed once senotherapeutics reach people. The study included 60 postmenopausal women and tested an intermittent senolytic regimen over 20 weeks as an approach to age-related bone loss.
The primary endpoint was bone resorption. The trial reported no observed change in that endpoint in the senolytic cohort. At two and four weeks, the study did find a statistically significant increase in a marker of bone formation, which was a secondary endpoint. A post-hoc analysis also found a different response among women in the highest tertile for T-cell p16 mRNA: in that subgroup, the bone-formation marker increased, the bone-resorption marker decreased and bone mineral density at the radius improved.
Those observations supply narrower signals, not a general treatment conclusion. The primary endpoint did not change, while the secondary and post-hoc findings involved different measures and a selected subgroup. They do not establish fracture prevention or validate a method for choosing patients for senolytic therapy.
No serious adverse events were described in the treated cohort during the study described by the review. That finding is useful for interpreting this trial, but it does not establish long-term safety across patients, doses, tissues or diseases. The clinical question is therefore becoming more specific: which people have a disease-relevant senescent-cell burden, which tissue should be targeted, what outcome should change and how long should benefit and unwanted effects be followed?
What targeted senescence treatment could mean for patients
The practical possibility is a treatment that helps patients retain tissue function or slow damage from an age-related disease by reducing senescence that is contributing to that disease. Reaching that point would require more than removing cells in an animal model or detecting an associated protein in blood.
Researchers would need to identify the relevant cellular population, demonstrate that the intervention changes disease biology at the right site and show a meaningful clinical outcome while controlling unwanted effects. Better tests for local senescence and treatment response could make that path more reliable, but they are means to the patient benefit rather than the benefit itself.
From the review’s November 2024 perspective, targeted senescence treatment remains a conditional prospect. Its value will depend on turning a biologically diverse cell state into a safely measured and appropriately treated contributor to disease, without mistaking every senescent cell for a harmful one.
Key Sources
- Primary source: Emerging insights in senescence: pathways from preclinical models to therapeutic innovations.
DOI or Journal Page · Review or Meta-Analysis · 2024-11-22
Keep exploring