Abstract

MLL-rearranged (MLLr) leukemia is characterized by a poor prognosis. Depending on the cell of origin, it differs in the aggressiveness and therapy response. For instance, in adults, volasertib blocking Polo-like kinase 1 (PLK-1) exhibited limited success. Otherwise, PLK-1 characterizes an infant MLLr signature, indicating potential sensitivity. By using our CRISPR/Cas9 MLLr model in CD34+ cells from human cord blood (huCB) and bone marrow (huBM) mimicking the infant and adult patient diseases, we were able to shed light on this phenomenon. The PLK-1 mRNA level was significantly increased in our huCB compared to the huBM model, which was underpinned by analyzing infant and adult MLLr leukemia patients. Importantly, the expression levels correlated with a functional response. Volasertib induced a significant dose-dependent decrease in proliferation and cell cycle arrest, most pronounced in the infant model. Mechanistically, upon volasertib treatment, we uncovered negative feedback only in the huBM model by compensatory upregulation of PLK-1 and related genes like AURKA involved in mitosis. Importantly, the poor response could be overcome by a combinatorial strategy with alisertib, an Aurora kinase A inhibitor. Our study emphasizes the importance of considering the cell of origin in therapeutic decision-making and provides the rationale for evaluating volasertib and alisertib in MLLr leukemia.