A naturally occurring shorter form of the p53 protein reduced an inflammatory signal called CXCL10 in several experiments involving progeria-model mice, naturally aged mice and cultured cells. The researchers also found an inverse association between the two molecules in human spleen data.
The protein variant, Δ133p53α, is a p53 isoform. Isoforms are different versions of a protein produced from the same gene. In this case, Δ133p53α lacks the first 132 amino acids found at the front end of full-length p53, giving it different effects on some p53-controlled pathways.
The mouse experiments suggest that Δ133p53α can restrain CXCL10 expression in specific tissues. The human evidence is much weaker: it comes from an observational analysis of 227 spleen samples and cannot show that one molecule controls the other in people. The study is a bioRxiv preprint and has not completed peer review.
Evidence at a glance
- Study type: Mechanistic experiments in transgenic mice and cultured cells, plus observational analysis of human tissue gene-expression data.
- Models: Progeria-model mice, naturally aged wild-type mice, mouse embryonic fibroblasts, fibroblasts from people with Hutchinson-Gilford progeria syndrome and 227 human spleen samples from GTEx.
- Evidence level: Early Research. This is a preprint.
- Main finding: Induced expression of human Δ133p53α lowered Cxcl10 in the spleen and brain of naturally aged mice and in several tissues of progeria-model mice. Human spleen samples showed an inverse association between Δ133p53α and CXCL10 RNA levels.
- Main caution: Mouse groups were small, the human analysis was correlative and the study did not test a therapy or show that lowering CXCL10 improves healthy aging.
What p53 normally does
Full-length p53 is best known as a tumor-suppressor protein. When cells experience DNA damage or other serious stress, p53 can pause the cell cycle, promote DNA repair, push a cell into senescence or trigger programmed cell death. These responses help prevent damaged cells from multiplying into tumors.
That protective role also creates biological trade-offs. Persistent p53 activity can contribute to long-lasting growth arrest and senescence in stressed tissues. Senescent cells no longer divide, but they remain metabolically active and can release inflammatory molecules known collectively as the senescence-associated secretory phenotype, or SASP.
Senescence is not simply harmful. It can suppress cancer, assist wound healing and shape development. Problems may arise when senescent cells accumulate or inflammatory signaling persists. Any attempt to alter p53 biology therefore has to balance tissue repair and inflammation against cancer protection.
How Δ133p53α differs
The human TP53 gene can generate multiple p53 isoforms. Δ133p53α begins from an internal start site and lacks much of the full-length protein’s front end. Prior work has found that it can inhibit selected p53-driven senescence programs while preserving or enhancing some DNA-repair activity.
Δ133p53α is naturally found in humans and other primates, not in ordinary mice. To study it in vivo, the researchers used a transgenic mouse strain engineered to express the human isoform after tamoxifen activated a Cre-based genetic switch. That system is useful for asking mechanistic questions, but it is not equivalent to a drug and does not reproduce the natural regulation of the isoform in people.
Why the researchers focused on CXCL10
CXCL10 is a chemokine, a small signaling protein that helps direct immune-cell movement and coordinate inflammation. It plays useful roles in acute responses to infection and tissue injury. Persistent production, however, can contribute to chronic inflammatory environments.
The team initially used a 32-plex assay to measure many cytokines and chemokines in serum from 15-week-old progeria-model mice. Each group contained four mice. As in the team’s earlier work, Δ133p53α expression reduced IL-6. It also lowered CXCL1, IL-1α and CXCL10, leading the researchers to examine CXCL10 in greater detail.
What changed in progeria-model mice
The progeria model carried one copy of the Lmna G609G mutation, which produces progerin and causes accelerated aging-like disease in mice. In both 15-week-old groups of four mice and 10-month-old groups of five mice, animals expressing Δ133p53α had lower Cxcl10 RNA in the spleen, brain and liver than non-expressing progeria controls. The same clear pattern was not seen in the lung.
RNA in situ hybridization, a method that visualizes specific RNA molecules in tissue sections, supported the spleen result in 10-month-old animals. This showed that the signal was present in the tissue rather than arising only from a whole-organ average.
Progeria models are valuable because they magnify certain forms of DNA damage, nuclear stress, senescence and inflammation. They do not reproduce ordinary human aging. A pathway that changes accelerated disease in a mouse may behave differently across the slower and more varied process of aging in people.
Results in naturally aged mice
The researchers also examined two-year-old wild-type mice without the progeria mutation. In groups of four animals, induced Δ133p53α expression significantly reduced Cxcl10 RNA in the spleen and brain. The liver and lung did not show the same response.
This experiment makes the study more relevant to ordinary aging biology than a progeria model alone. It remains a small, engineered mouse experiment. The researchers measured molecular expression, not whether the mice became stronger, cognitively healthier or longer lived because CXCL10 fell.
Cell experiments suggest a direct regulatory effect
In mouse embryonic fibroblasts, turning on Δ133p53α for five days reduced Cxcl10 RNA, including in cells engineered to express progerin. A similar five-day experiment in two fibroblast lines derived from people with progeria also reduced CXCL10 RNA.
The short time frame argues against the mouse tissue results being only a long-term adaptation to lifelong transgene expression. It is consistent with a more direct regulatory role. The authors note that full-length p53 can activate CXCL10 transcription and propose that Δ133p53α may inhibit that activity. The exact molecular steps were not demonstrated in this study.
What the human spleen data add
The researchers analyzed RNA sequencing data from 227 human spleen samples in the Genotype-Tissue Expression project, known as GTEx. Samples with higher estimated Δ133p53α expression tended to have lower CXCL10 expression.
This inverse relationship gives the mouse and cell findings some human relevance, but it is observational. The samples came from different donors, and the analysis cannot establish direction, exclude confounding or show that changing Δ133p53α would lower CXCL10 in a living person. It also does not demonstrate a health benefit.
Why p53 manipulation demands caution
A pathway that reduces senescence-associated inflammation may look attractive as a future therapeutic target. p53 is not an isolated inflammation switch. It sits at the center of cancer suppression, DNA-damage responses, cell-cycle control, apoptosis and senescence.
Broadly weakening p53 activity could allow damaged cells to survive or divide. Increasing Δ133p53α could also have different effects depending on the cell type, stress level, dose and duration. The authors’ transgenic system was tightly controlled in research animals. Translating that biology would require much more selective tools and long-term cancer surveillance.
Limitations
The animal groups were small, generally four or five mice, which makes estimates less precise and increases the need for independent replication. The experiments used transgenic expression of a human isoform in mice rather than a clinically realistic intervention.
Many tissue results measured Cxcl10 RNA rather than CXCL10 protein or downstream immune-cell behavior. The study did not directly show that reduced CXCL10 caused any previously observed improvement in pathology or survival. Tissue-specific differences, including the lack of a consistent lung response, remain unexplained.
The human spleen analysis was correlative and based on postmortem tissue data. It did not manipulate the pathway, follow people over time or test clinical outcomes. Finally, the work is a preprint and has not undergone journal peer review.
What the research does not show
The study does not establish Δ133p53α as an anti-aging treatment. It does not show that increasing the isoform is safe in people, that blocking CXCL10 improves human healthspan or that the pathway extends lifespan during normal aging.
It also does not prove that CXCL10 is the key mediator of Δ133p53α’s effects. Several inflammatory molecules changed, and p53 isoforms influence multiple cellular programs. A lower biomarker is not automatically evidence of a better clinical outcome.
What researchers need to demonstrate next
Independent laboratories should replicate the results with larger groups and measure CXCL10 protein, immune-cell recruitment, tissue function and long-term outcomes. Directly raising or blocking CXCL10 would help determine whether it is necessary for the benefits attributed to Δ133p53α.
Mechanistic studies could test whether full-length p53 binds and activates the CXCL10 regulatory region and how Δ133p53α changes that interaction. Cell-type-specific experiments would show where the pathway matters and where it could be risky.
Any translational program would need a precise way to alter the pathway, pharmacology that can be stopped or reversed, and long follow-up for cancer and other adverse effects. In human studies, longitudinal tissue or blood data would be more informative than a single postmortem correlation.
The Lifespan Brief assessment
The study identifies CXCL10 as a plausible component of the anti-inflammatory biology associated with Δ133p53α. Its most convincing evidence is mechanistic and preclinical: small, controlled experiments in engineered mice and cultured cells produced a consistent tissue-specific signal.
The human evidence is supportive but observational. This work is best viewed as a pathway map for future research, not a treatment claim. The central challenge is to learn whether CXCL10 meaningfully drives pathology and whether any intervention can preserve p53’s cancer-protective functions.
The findings also connect with broader work on progeria, chromatin disruption and senescence. A recent Lifespan Brief report examined a different peptide strategy in a progeria mouse model. Both studies offer mechanistic clues, but neither establishes a therapy for normal human aging.
Sources
Yamada and colleagues, The senescence-inhibitory p53 isoform Δ133p53α represses the proinflammatory chemokine CXCL10 in progeria model mice and naturally aged mice, bioRxiv, version 3 posted June 22, 2026. Preprint, not peer reviewed.
Fujita and colleagues, p53 isoforms Δ133p53 and p53β are endogenous regulators of replicative cellular senescence, Nature Cell Biology, 2009.
Horikawa and colleagues, p53 isoforms regulate premature aging in human cells, Blood, 2018.