Ran Tao spent years chasing a ghost in the genome — a faint signal buried in layers of genetic code that, once uncovered, might change the fate of children with aggressive brain tumors. Today, that signal has a name: KDM2B. In a breakthrough study, scientists at St. Jude Children's Research Hospital and Hopp Children's Cancer Center Heidelberg (KiTZ) discovered that two high-risk types of medulloblastoma — Group 3 and Group 4 — depend on this single protein to grow. These tumors, which strike mostly in childhood, are among the most common malignant brain cancers in kids, yet they still lack targeted treatments. Most children endure grueling therapies like radiation and chemotherapy, which can cause lifelong side effects. Finding a precise target like KDM2B could one day change that.
Medulloblastoma is divided into four main subgroups, but Group 3 and Group 4 have long posed a challenge because they don’t respond to the same therapies as the others. Researchers suspected that changes in how genes are regulated — not in the genes themselves — might be driving these tumors. These are called epigenetic changes, where chemical tags on DNA or its packaging alter gene activity without changing the genetic code. To find the culprits, the team analyzed 52 patient tumor samples, the largest such study of its kind, profiling six different histone modifications across all subgroups. This created a detailed map of how epigenetics shapes each tumor type.
That map led them straight to KDM2B. The protein was highly active in Group 3 and Group 4 tumors, and when scientists removed or degraded it in lab models, tumor growth slowed and survival improved. Further investigation revealed that KDM2B doesn’t work alone — it recruits a team of proteins called Polycomb repressive complexes, which help silence genes that would otherwise stop cancer from growing. By blocking KDM2B or its partners, researchers may be able to disrupt this dangerous network.
"We found that select medulloblastoma groups require KDM2B to grow using preclinical models," said Paul Northcott, Ph.D., who led the study. "This gives us a potential future therapeutic opportunity that may selectively suppress the growth of Group 3 and Group 4 medulloblastomas." The discovery, years in the making, opens a new path toward smarter, kinder treatments for children who need them most. While still in early stages, the findings published in Nature Genetics offer real hope: that one day, a single molecule could help dismantle a deadly cancer from within.