Childhood Brain Tumor Study Exposes KDM2B as a Treatable Weak Spot in High-Risk Medulloblastoma

Scientists at St. Jude Children’s Research Hospital and the Hopp Children’s Cancer Center Heidelberg (KiTZ) have identified a shared molecular vulnerability in two of the most aggressive subtypes of medulloblastoma, the most common malignant brain tumor of childhood. The protein KDM2B, an epigenetic regulator, turns out to be essential for the growth of Group 3 and Group 4 medulloblastomas, the subgroups that carry the highest risk and currently have no targeted therapies. When the researchers removed the gene encoding KDM2B or degraded the protein itself in preclinical models, tumor growth slowed and survival was prolonged. The work, published on September 24, 2026, in Nature Genetics, offers the first realistic path toward precision medicine for children whose tumors have long been treated only with blunt, intensive approaches.

Medulloblastoma arises in the cerebellum and comprises four major molecular subgroups: WNT, SHH, Group 3, and Group 4. While WNT tumors have an excellent prognosis and SHH tumors can sometimes be addressed with inhibitors of the sonic hedgehog pathway, Group 3 and Group 4 remain stubbornly resistant to targeted approaches. Children with these tumors typically receive maximal surgery, craniospinal irradiation, and multi-agent chemotherapy. Although survival rates have improved over decades, the intensity of treatment leaves many survivors with lasting cognitive, endocrine, and developmental damage. This therapeutic gap is precisely what motivated the St. Jude and KiTZ collaboration: if no drugs exist for these subgroups, the field must first discover what the tumors cannot live without.

The biological logic behind the search lies in epigenetics. Epigenetic regulation refers to the layer of control that determines which genes are switched on or off without altering the underlying DNA sequence. The principal players are chemical marks deposited on histones, the protein spools around which DNA is wound. Distinct histone modifications, such as methylation and acetylation at specific amino acid positions, act as signposts that recruit or repel the molecular machinery of transcription. Because epigenetic deregulation that disrupts normal brain development has long been suspected of driving medulloblastoma formation, mapping the exact histone modification landscape of patient tumors promised to reveal where that control breaks down, and which proteins might be holding the broken system together.

Delivering that map required an effort of unusual scale for a rare pediatric cancer. The team performed the largest analysis of its kind in medulloblastoma, profiling 52 primary patient samples drawn from all four major subgroups, with multiple samples per subgroup to ensure that subgroup-specific signals could be distinguished from individual tumor quirks. They measured six distinct histone post-translational modifications and used chromatin state, the functional packaging configuration of the genome, as a readout of epigenetic activity. The resulting dataset, made publicly available through a web portal so that other investigators can mine it, constitutes a comprehensive community resource. The researchers deliberately designed the resource to help the field move beyond cataloging mutations toward identifying functional dependencies.

The search for strong subgroup-specific signals quickly paid off. The team detected a recurrent, aberrant epigenetic signature shared by Group 3 and Group 4 tumors, an unexpected convergence given that these subgroups differ in their genomic features and presumed cells of origin. When the researchers asked which chromatin-modifying protein could account for the signature, KDM2B emerged as the leading candidate. First author Ran Tao, PhD, of the St. Jude Department of Developmental Neurobiology described the progression: after identifying the signature, the team spent years investigating how KDM2B controls tumor growth and working through the underlying molecular mechanisms. KDM2B is highly expressed in Group 3 and Group 4 tumors, and its abundance marks it as a defining feature of these high-risk disease categories.

Functionally, KDM2B proved to be a genuine dependency rather than a passive passenger. In preclinical models, genetic removal of the KDM2B gene or pharmacologically oriented degradation of the protein suppressed tumor growth and extended survival. Corresponding author Paul Northcott, PhD, director of the St. Jude Center of Excellence in Neuro-Oncology Sciences and a member of the Department of Developmental Neurobiology, summarized the finding: select medulloblastoma groups require KDM2B to grow, which gives the field a potential future therapeutic opportunity to selectively suppress the growth of Group 3 and Group 4 tumors while sparing normal tissue. The selectivity is the crucial point, because an ideal pediatric target should be dispensable in healthy cells but indispensable in the tumor.

The project itself has a long history that illustrates how cancer genomics matures over time. Northcott traced its origins to his postdoctoral years, when early sequencing of medulloblastoma genomes revealed an enrichment of alterations in chromatin-modifying genes, a signal that at the time had no obvious therapeutic exploitation. Almost fifteen years later, with modern epigenomic profiling tools and the infrastructure to study dependencies, that old signal could finally be pursued to its mechanistic roots. The trajectory underscores a broader lesson in pediatric oncology: discoveries from large-scale sequencing projects often take years, sometimes decades, before they translate into actionable biology, and sustained investment in data generation and model systems is what makes such translation possible.

KDM2B itself is not the end of the story, and perhaps not even the easiest drug target. The mechanistic studies showed that KDM2B exerts its growth-promoting influence by recruiting Polycomb repressive complexes, a family of protein assemblies that modify chromatin to silence genes. Recognizing this downstream dependency, the team looked for additional vulnerabilities among the complexes themselves. Data from The Pediatric Cancer Dependencies Accelerator, a multi-institutional initiative involving the Broad Institute, Dana-Farber Cancer Institute, and St. Jude, indicated that removing these Polycomb complexes in medulloblastoma cell lines also slows cancer growth. This downstream angle matters clinically because drugging a complex can be more tractable than drugging a single regulatory protein.

Two Polycomb-related findings stand out for their therapeutic implications. First, PRC2, one of the Polycomb repressive complexes, is already being investigated as a drug target in other pediatric cancers, which lends external validation to the strategy and suggests that existing medicinal chemistry and clinical experience could be adapted for medulloblastoma trials. Second, the study identified PRC1.1 as a previously unrecognized potential therapeutic target for Group 3 and Group 4 medulloblastoma, adding a novel candidate to a therapeutic landscape that previously held none. Together, these findings reframe the problem: targeting KDM2B directly, or disrupting the Polycomb pathways it recruits, provides at least three distinct intervention points against tumors that have historically been treated with one-size-fits-all regimens.

The researchers emphasize that the public dataset is as much a product of the study as the KDM2B discovery itself. By characterizing the medulloblastoma chromatin landscape in unprecedented detail and making it freely explorable, the team hopes that other laboratories will mine the resource for additional subgroup-specific signatures and dependencies. Northcott framed the ambition plainly: the characterization identified a potential vulnerability in subgroups that still lack targeted therapy, and he hopes others will explore the data to make further discoveries and, collectively, find new therapeutic targets for subgroups with poor prognosis. For the children diagnosed with Group 3 and Group 4 medulloblastoma each year, the study converts fifteen years of accumulated genomic insight into a concrete biological target, a set of downstream pathway candidates, and a roadmap for developing the selective treatments that this high-risk population has long awaited.

Subject of Research: KDM2B dependency in Group 3 and Group 4 medulloblastoma epigenetics

Article Title: KDM2B revealed as a selective dependency in high-risk medulloblastoma subgroups

Article References: KDM2B revealed as a selective dependency in high-risk medulloblastoma subgroups. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: medulloblastoma, KDM2B, epigenetics, histone modifications, Polycomb repressive complexes, Group 3 medulloblastoma, Group 4 medulloblastoma, pediatric brain tumor, targeted therapy, chromatin regulation, PRC2, preclinical models