(MEMPHIS, Tenn. – March 25, 2026) Researchers have uncovered the long-sought explanation for how the most common cause of pediatric ependymoma, ZFTA-RELA fusions, leads to disease. Scientists at St. Jude Children's Research Hospital and Baylor College of Medicine discovered that the ZFTA-RELA fusion oncoprotein exploits open regions of the genome which control growth, trapping immature brain cells in a perpetual state of immature development. The work, published today in Nature, highlights a previously unappreciated way in which fusion oncoproteins contribute to cancer.
Cancers are often caused by abnormal proteins that rewire a cell's genomic structure, forcing open gene programs dedicated to cell growth. However, when scientists reproduced the most frequently seen alteration in ependymoma, the fusion of ZFTA and RELA, this did not happen as expected.
"When we expressed the fusion protein, we observed very few changes in the patterns of chromatin accessibility," said co-corresponding author Stephen Mack, PhD , St. Jude Department of Developmental Neurobiology . "We concluded the programs are largely set up, and when the fusion event happens, it simply accesses these regulatory sites and keeps them on."
ZFTA-RELA helps shape ependymoma cell diversity
Through their studies, the researchers realized that ZFTA-RELA is replicating the activity of PLAG/L, a family of proteins that control access to cell development programs. PLAG/L proteins bind a specific DNA sequence to switch these cell development programs on. This process is required for immature cell development but typically shuts down when development is complete. ZFTA/RELA binds the same DNA sequence as PLAG/L to keep cell development programs open, trapping the cells in developmental limbo.
"If we better understood how normal cells shut down PLAG/L chromatin accessibility, it might provide ways to impair ZFTA-RELA oncogenic activity," said Alisha Kardian, St. Jude Department of Developmental Neurobiology and Baylor College of Medicine graduate student.
Additionally, the researchers found that ZFTA/RELA ependymoma cells largely stem from a few dominant ancestors despite the abnormal activation of different developmental gene programs. This implies that one or a few "winner" cells take over, but these cells can access multiple developmental programs. Cell dominance appeared to coincide with a ZFTA-RELA expression sweet spot.
"Too little fusion doesn't sufficiently activate programs; too much is toxic," said Mack. "But this highlights why missing even a few cells during therapy can lead to relapse."
Currently, treatment for pediatric ependymoma relies on surgery and radiation. But these findings may spark novel therapeutic strategies to test, such as forcing cells past immature developmental states to resolve the ZFTA-RELA roadblock.
"Ependymomas resist conventional chemotherapy," said co-corresponding author Kelsey Bertrand, MD , St. Jude Department of Oncology . "We believe these findings will help better understand how this happens."
Authors and funding
The study's co-corresponding author is Benjamin Deneen, Baylor College of Medicine. The study's co-first author is Hua Sun, St. Jude. The study's other authors are Taylor Bugbee, Sanya Mehta, Sophie Cochiolo, Blake Holcomb and Kaitlin Budd, St. Jude Graduate School of Biomedical Sciences; Siri Ippagunta, Nicholas Laboe, Srinidhi Varadharajan, Erik Emanus, Tuyu Zheng, Jon Connelly, Yong-Dong Wang, Kimberley Lowe, Rakesh Pathak, Amanda Bland, Amir Arabzade, Gabriele Kembuan, Tristen Wright, Emma Caesar, Maxwell Park, Thomas Merchant, Adam Durbin, Alfonso Lavado, Suzanne Baker, David Ellison, and Shondra Pruett-Miller, St. Jude; Kwanha Yu, Hsiao-Chi Chen and Riley Deneen, Baylor College of Medicine; Amelia Hancock, David Gee and Joel Murdoch, University of Bath; Jiangshan Zhan and Hyoung-Kyoung Lee, Baylor College of Medicine and Texas Children's Hospital; Yi Xiao and Samuel McBrayer, UT Southwestern; Jun Qi, Dana-Farber Cancer Institute; Li Wang, University of California San Francisco; Andrew Donson and Nicholas Foreman, University of Colorado Anschutz Medical Campus; and Sameer Agnihotri, University of Pittsburg and Children's Hospital of Pittsburg.
The study was supported by the National Cancer Institute (P30CA021765, P01CA298963-01, R01NS128184, R01CA280203, R01CA284455, U01CA281823, U01CA217842, P01CA096832, P30CA021765), the Department of Defense (CA220510 and CA220247), the National Institutes of Health (R35NS132230, R01NS124093, R01CA223388 and F31CA265285), the National Brain Tumor Society, CERN Foundation, Alex's Lemonade Stand Foundation 'A' Award, the Helis Medical Research Foundation, the National Institute of Neurological Disorders and Stroke (2R01NS110859 and R01NS126287), the Wallace Endowment, the Petrello family, the National Multiple Sclerosis Society (FG240743793) and the American Lebanese Syrian Associated Charities (ALSAC), the fundraising and awareness organization of St. Jude.
