Functional, structural, and molecular characterizations of the leukemogenic driver MEF2D-HNRNPUL1 fusion

Recurrent MEF2D fusions with poor prognosis have been identified in B-cell precursor ALL (BCP-ALL). The molecular mechanisms underlying the pathogenic function of MEF2D fusions are poorly understood. Here, we show that MEF2D-HNRNPUL1 (MH) knock-in mice developed a progressive disease from impaired B-cell development at the pre–pro-B stage to pre-leukemia over 10 to 12 months. When cooperating with NRAS(G12D), MH drove an outbreak of BCP-ALL, with a more aggressive phenotype than the NRAS(G12D)-induced leukemia. RNA-sequencing identified key networks involved in disease mechanisms. In chromatin immunoprecipitation–sequencing experiments, MH acquired increased chromatin-binding ability, mostly through MEF2D-responsive element (MRE) motifs in target genes, compared with wild-type MEF2D. Using X-ray crystallography, the MEF2D-MRE complex was characterized in atomic resolution, whereas disrupting the MH-DNA interaction alleviated the aberrant target gene expression and the B-cell differentiation arrest. The C-terminal moiety (HNRNPUL1 part) of MH was proven to contribute to the fusion protein’s trans-regulatory activity, cofactor recruitment, and homodimerization. Furthermore, targeting MH-driven transactivation of the HDAC family by using the histone deacetylase inhibitor panobinostat in combination with chemotherapy improved the overall survival of MH/NRAS(G12D) BCP-ALL mice. Altogether, these results not only highlight MH as an important driver in leukemogenesis but also provoke targeted intervention against BCP-ALL with MEF2D fusions.