Novel strategy Against AML by Rescuing Loss of Phase Separation of Pathogenic NPM1c

Runmeng Liu, BE (Advisor: Xiaotian Zhang, PhD)

          Nucleophosmin (NPM1) is a multifunctional protein predominantly localized to the nucleolus. In approximately 30% of acute myeloid leukemia (AML) cases, NPM1 is mutated, resulting in the aberrant cytoplasmic localization of the mutant protein (NPM1c). These mutations generate a novel C-terminal nuclear export signal (NES), enabling interaction with XPO1 and promoting the relocalization of NPM1c condensates to genomic loci such as the HOX gene clusters and MEIS1, thereby driving transcriptional hijacking. However, the mechanisms underlying NPM1c condensate formation, transcriptional reprogramming, and maintenance of the undifferentiated leukemic state remain incompletely understood. In this study, we employed OCI-AML3 NPM1c degron 2 cells as a model system. By expressing NPM1c and its variants under an inducible system while degrading endogenous NPM1c, we systematically dissected the structural determinants governing NPM1c function. We demonstrate that the oncogenic condensate formation is primarily driven by the gain of the C-terminal NES resulting from the frameshift mutation, rather than the loss of the nucleolar localization signal (NoLS). Importantly, the leukemogenic function of NPM1c condensates is independent of their subcellular localization but critically depends on the integrity of the full-length C-terminal NES. Furthermore, we show that reintroduction of the RNA-binding region of NPM1 partially suppresses the oncogenic function conferred by the NPM1c C-terminal NES. This finding suggests that the loss of RNA-binding capacity contributes to the maintenance of the undifferentiated state in AML cells, potentially through the disruption of tumor-suppressive mechanisms. Nevertheless, the NES-driven gain-of-function appears to play a dominant role in oncogenic reprogramming. Finally, mutational disruption of key nucleic acid–interacting residues within the C-terminal helical domain of NPM1c, particularly those affecting side-chain polarity and interaction with the nucleic acid phosphate backbone, significantly impairs the ability of NPM1c condensates to sustain the undifferentiated state of AML cells. These results suggest that specific acidic residues, such as aspartate, may contribute to transcriptional hijacking through nucleic acid interactions within NPM1c condensates. In summary, our study identifies the C-terminal NES and nucleic acid-binding capacity of NPM1c as critical determinants of condensate formation, transcriptional reprogramming, and maintenance of the undifferentiated leukemic state. These findings provide mechanistic insights into NPM1c-driven leukemogenesis and highlight potential therapeutic targets for AML.

Advisory Committee:

• Xiaotian Zhang, PhD, Chair
• Xiaodong Cheng, PhD
• Bruno Di Stefano, PhD
• Andrew Dunbar, MD
• Jian Hu, PhD