Medulloblastoma is the most common malignant brain tumor in children. Despite management combining surgical resection, chemotherapy and craniospinal irradiation, the prognosis remains highly variable: approximately one third of patients succumb to the disease. Molecular classifications have made it possible to distinguish four main subgroups — Sonic Hedgehog, Wingless/Int-1, Group 3 and Group 4 — each characterized by distinct transcriptional profiles, genomic alterations and clinical courses. While the (epi)genomic and transcriptomic characterization of these tumors is now well advanced, proteomic analysis has only recently emerged and metabolomic studies are still in their infancy. Yet a notable subset of Group 3 tumors, among the most lethal, is distinguished by aberrant expression of the MYC oncogene, with no satisfactory therapeutic option.
To address this gap, the authors assembled a large multi-omics dataset integrating five levels of information — CpG methylome, transcriptome, proteome, phosphoproteome and metabolome — derived from 384 primary patient samples, together with their clinical metadata. Unsupervised integration of these layers revealed heterogeneity in lipid metabolism across proteomic subtypes, particularly pronounced within Group 3. The analyses were then extended through functional experiments carried out in vitro and in mouse models, relying on cell lines, pharmacological inhibition of several lipid metabolism enzymes, and various imaging and cellular metabolism approaches.
This work reveals that the MYC-FASN-SCD axis sustains lipid biosynthesis in these tumors. However, inhibition of this pathway triggers in vivo a compensatory escape mechanism based on the uptake of exogenous fatty acids, which limits its efficacy as a single target. Unexpectedly, the team demonstrates that MYC drives lipid storage and thereby creates a unique dependence on communication between lipid droplets and mitochondria, required for tumor maintenance in vivo. This dependence constitutes a vulnerability located downstream of MYC.
By mapping the metabolic landscape of medulloblastoma on a large scale, this study thus identifies a targetable vulnerability specific to the MYC-driven Group 3 subtype. The authors conclude that lipid storage and the crosstalk between lipid droplets and mitochondria represent a promising therapeutic avenue against these tumors that currently lack effective treatment. Beyond this finding, the full dataset generated, available through public repositories, provides an integrated resource for further exploration of the biology of this disease.