Radiation Turns Childhood Brain Tumor Cells Senescent—and Senolytic Drugs Can Kill Them

Medulloblastoma is the most common malignant brain tumor of childhood, and for decades the backbone of its treatment has remained largely unchanged: maximal surgical removal, followed by craniospinal irradiation and chemotherapy. That regimen saves many lives, but it comes at a steep price. Radiation aimed at the entire brain and spine can trigger endocrine dysfunction, inflammation, and lasting neurocognitive impairment in young patients whose developing nervous systems are especially vulnerable. A new study published in Cell Death Discovery now suggests a surprising way to soften this trade-off—by exploiting a cellular state that radiation itself creates inside the tumor.

Researchers at Heinrich-Heine-University Düsseldorf, led by Dennis Sohn and colleagues, report that pediatric medulloblastoma cells carrying functional, wild-type p53 do not simply die when bombarded with gamma irradiation. Instead, many of them enter senescence, a terminal and stable arrest of the cell cycle. These zombie-like cells stop dividing but refuse to die, lingering in the tissue for weeks. Crucially, the team showed that these irradiated, senescent tumor cells can then be selectively wiped out with senolytic drugs—compounds designed to eliminate senescent cells. The finding opens a potential two-step strategy: radiation first drives susceptible tumor cells into senescence, and senolytics then clear them away.

The study compared two SHH-activated medulloblastoma cell lines with fundamentally different genetic backgrounds. ONS76 cells carry wild-type p53, the master tumor suppressor that decides a cell’s fate after catastrophic DNA damage. UW228-3 cells carry a mutated, dysfunctional p53. When both lines were exposed to a 20 Gy dose of gamma irradiation, their responses diverged dramatically. UW228-3 cells detached from the culture dish and degraded completely within three to seven days. ONS76 cells, by contrast, remained attached for more than fifty days, flattening out and swelling enormously in both cell body and nucleus—classic hallmarks of senescence.

The molecular evidence followed the morphology. Irradiated ONS76 cells showed a dose-dependent stabilization of p53 and a strong increase in p21, the cyclin-dependent kinase inhibitor encoded by the CDKN1A gene that enforces the senescence arrest. They also accumulated senescence-associated beta-galactosidase activity, a lysosomal enzyme widely used as a senescence marker, and built up lipofuscin, the pigment-like cellular debris characteristic of aged or senescent cells. When the researchers knocked down p53 or p21 using siRNAs, the senescence program collapsed: far fewer cells became senescent and substantially more died instead. This established that radiation-induced senescence in these cells is strictly p53/p21-dependent.

The p53-mutated UW228-3 cells told the opposite story. They showed only a faint, transient rise in beta-galactosidase activity and never developed senescent morphology. Instead, measurements of lactate dehydrogenase release—a marker of plasma membrane destruction—and of DEVDase-like caspase activity revealed that they died by apoptosis, the controlled self-destruction pathway. A pan-caspase inhibitor, q-VD-OPh, completely abolished this cell death while leaving senescence untouched in the wild-type line. The team also employed a recently developed fluorescent lipofuscin binder, GLF16, alongside the amine-reactive viability dye Zombie-violet, allowing simultaneous flow cytometric quantification of senescent and dead cells with sharper discrimination than traditional staining.

The researchers then turned to c-Myc, a proto-oncogene that serves as a crucial negative prognostic factor in medulloblastoma and is known to antagonize p21. Using engineered ONS76 and UW228-3 lines that constitutively overexpress c-Myc, they found that the oncogene tipped the balance away from senescence. In irradiated ONS76/c-Myc cells, p21 protein levels dropped—though p21 mRNA levels did not, indicating that c-Myc interferes with p21 protein production or stability rather than its transcription. Fewer cells adopted senescent morphology, and more cells died, a death that q-VD-OPh could block even though it was not accompanied by elevated caspase activity. Notably, this c-Myc effect operated independently of p53 status. The senescent population in c-Myc-overexpressing cells also proved far more resistant to senolytic treatment, underscoring how oncogene dosage can shape therapeutic vulnerability.

The therapeutic payoff came in the final set of experiments. Seven days after irradiation, when the ONS76 cultures had become fully senescent, the researchers treated them with four senolytic compounds: the BCL-2 family inhibitors Navitoclax and Venetoclax, and the PI3K inhibitors PX-866 and BAY 80-6946 (Copanlisib), the latter already FDA-approved. All four killed the senescent cells in a dose-dependent manner. Navitoclax, however, also eliminated proliferating cells at the highest concentration tested, 30 µM—a reminder of the well-known toxicity problem that has complicated its clinical use, since healthy non-senescent cells are targeted at higher doses. The PI3K inhibitors, by contrast, showed a cleaner specificity for the senescent population.

The clinical logic behind this approach is compelling. Medulloblastoma is unusual among cancers in that relatively few tumors harbor p53 mutations, and in the SHH-activated subtype, p53 status actually defines two distinct risk groups: p53-mutated tumors carry a five-year overall survival of roughly 41 percent and are classified as high risk, while p53 wild-type tumors fare far better at about 81 percent. The new findings suggest that these two groups respond to radiotherapy in mechanistically different ways—senescence in the wild-type setting, apoptosis in the mutated one—a distinction invisible to standard proliferation-based cytotoxicity assays. Senolytic combination therapy would therefore primarily apply to the intermediate-risk, p53 wild-type group, where it could potentially improve outcomes further while reducing the radiation dose needed for tumor control.

The broader context strengthens the case. Senescent cells are not inert; they secrete a cocktail of cytokines and chemokines known as the senescence-associated secretory phenotype, or SASP, which drives chronic tissue microinflammation and can support tumor relapse. Craniospinal irradiation is associated with accelerated brain aging in survivors, and studies in glioblastoma have already shown that eliminating therapy-induced senescent cells with Navitoclax attenuates recurrence. Mouse studies of whole-brain irradiation likewise suggest that senolytic clearance protects cognitive function and the blood-brain barrier. Because senescence can be triggered at lower radiation doses than outright cell death, pairing senolytics with craniospinal irradiation might one day allow clinicians to cure tumors with less radiation exposure while simultaneously mopping up the senescent cells that radiation leaves behind.

Important caveats remain. The entire study was conducted in cell lines rather than animal models or patients, and the exact nature of the c-Myc-driven, caspase-inhibitable cell death observed in the overexpressing cells is still unclear. The authors themselves note that it would be intriguing to test whether JQ1, a bromodomain inhibitor that suppresses c-Myc signaling, could push c-Myc-high, p53 wild-type medulloblastoma cells back into a senescence-permissive state and thereby restore their susceptibility to senolytics. Still, the work is the first to document radiation-induced senescence in medulloblastoma cells and to demonstrate that these cells can be eliminated with existing senolytic drugs. If the strategy survives translation into preclinical models, it could reshape how one of pediatric oncology’s most grueling treatments is delivered—turning the very cells that radiation strands between life and death into targets for a second, cleaner strike.

Subject of Research: Radiation-induced cellular senescence in pediatric medulloblastoma and its targeting by senolytic drugs

Article Title: p53 wild-type/low c-Myc-expressing ONS76 pediatric medulloblastoma cells become senescent after gamma-irradiation and can be eliminated by senolytic substances

Article References: Sonntag, J., Preugschas, R.-L., Jazmati, D., Qin, N., Neuwahl, J., Matuschek, C., Remke, M., Budach, W., Hörner-Rieber, J., & Sohn, D. (2026). p53 wild-type/low c-Myc-expressing ONS76 pediatric medulloblastoma cells become senescent after gamma-irradiation and can be eliminated by senolytic substances. Cell Death Discovery, 12(1), Article 410. https://doi.org/10.1038/s41420-026-03404-4

Image Credits: AI Generated

DOI: 10.1038/s41420-026-03404-4

Keywords: medulloblastoma, senescence, senolytics, p53, c-Myc, p21, gamma irradiation, craniospinal irradiation, Navitoclax, PI3K inhibitors, SASP, pediatric brain tumor