Abstract
Chronic myelogenous leukemia (CML) is a stem-cell driven malignancy caused by BCL/ABL, the oncoprotein derived from the Philadelphia chromosome 9/22 translocation. Although tyrosine kinase inhibitors (TKIs) are highly effective in managing chronic phase CML, challenges remain, including acquisition of TKI resistance, toxicities of long-term TKI administration and third-generation TKIs, and persistence of leukemia initiating cells (LICs). Reliable curative chemotherapy remains elusive. Alternative approaches to definitive treatment of CML would enhance current management.
We report here that the Polycomb group protein EZH2, a catalytic subunit of Polycomb repressive complex 2 (PRC2), is a potential target for curative chemotherapy for CML. EZH2 is overexpressed in LICs, compared to normal hematopoietic stem cells (HSCs), and is sustained by BCR/ABL signaling, as TKIs suppress EZH2 expression in CML cells. Removal of EZH2 in an engineered mouse model prevents CML initiation and development. Moreover, Ezh2 is also required for maintenance of established CML. In a tamoxifen-inducible EZH2 deletion mouse model, conditional inactivation of EZH2 eradicates existing disease induced by wild-type BCR/ABL or the gatekeeper T315I mutant, and leads to long-term survival. Importantly, EZH2 inactivation also eliminates phenotypic and functional LICs. Relevance of the genetic studies in mice is supported by shRNA knockdown and EZH2-inhibitor treatment of human CML cell lines and primary samples, which leads to cell cycle inhibition, apoptosis, and reduced colony formation. Effects of EZH2 loss in CML are context-specific, as EZH2 is dispensable for HSCs, in part due to decreased EZH1 expression in CML LICs compared to normal HSCs. EZH1 expression decreases even further in LICs upon the loss of EZH2. Thus, loss of EZH2 in CML LICs creates a scenario resembling complete inactivation of PRC2, which is essential for the maintenance of HSCs.
EZH2-dependence of CML LICs raises prospects for curative chemotherapy based on addition of EZH2 inhibitors to conventional TKI administration.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.