Targeting non-genetic kinase-dependent signaling pathways to prevent group 3 medulloblastoma tumor relapse

Abstract

Medulloblastoma (MB) is the most common pediatric brain tumor, accounting for about 25% of childhood brain tumor cases. Among its molecular subgroups, Group 3 (G3) is particularly aggressive, marked by high rates of metastasis at diagnosis and a devastating 30% recurrence rate that is nearly fatal because of its therapy-resistant nature. Although the transcriptional programs driving primary tumors are well characterized, the molecular mechanisms underlying the transition to therapy resistance remain poorly understood. This challenge is further complicated by the extreme scarcity of matched primary and recurrent patient specimens. To overcome this challenge, we established a clinically relevant experimental model in which G3 MB cells were treated with standard-of-care chemoradiotherapy (CRT), and living cells were isolated. We then mapped signaling adaptations in these CRT-surviving cell populations, discovering that the non-receptor tyrosine kinase SRC and its downstream effectors were significantly upregulated following CRT. Interestingly, this adaptive increase was exclusive to the G3 MB and was not observed in the SHH subgroup, Group 4 MB, normal human astrocytes, or neural stem cells. Functional validation through genetic ablation or pharmacological inhibition showed that SRC acts as a central survival anchor for these cells. Disrupting SRC affected the stemness circuitry, involving key factors such as SOX2, NOTCH1, OCT4, and NANOG, and further hindered both migratory and proliferative capabilities. The loss of progenitor identity was accompanied by increased expression of the neuronal marker TUBB3, indicating a forced shift toward terminal lineage differentiation. This fundamental change effectively stripped the cells of their plasticity, initiating dual-mode programmed cell death through both apoptotic and necroptotic pathways. Finally, repurposing the blood-brain-barrier-permeable SRC inhibitor, Saracatinib, in combination with standard CRT in patient-derived orthotopic xenograft models led to a significant reduction in tumor burden and an extension of overall survival, all without causing neurotoxicity. By uncovering a previously hidden mechanism of recurrence in G3 MB, our research provides a rationale for the clinical application of SRC inhibitors. This approach offers a transformative, neuro-sparing opportunity in pediatric oncology by specifically targeting the signaling pathways that enable the most aggressive tumors to survive, adapt, and recur.