Targeting the Epigenetic and Metabolic Vulnerabilities of Ezh2-Deficiency-Induced Myeloid Neoplasms

Myeloproliferative neoplasms (MPNs) are clonal, progressive hematopoietic malignancies characterized by alterations of multiple signaling and epigenetic pathways in hematopoietic stem cells (HSCs). The identification of driver oncogenic mutations in JAK2, CALR or MPL has transformed our knowledge of MPNs; however, patients with non-mutated JAK2, CALR and MPL (so-called 'triple-negative') have the highest incidence of leukemic transformation and the lowest overall survival (Rumi et al., 2014 Blood 124:1062-1069), indicating that other factors may contribute to MPNs. The molecular processes controlling MPN progression from chronic to acute leukemic transformation remain unknown. This is a major impediment for developing target-based therapeutics to selectively eliminate the mutant stem cells to prevent disease progression and/or relapse. Therefore, the challenge is twofold: 1) to identify and characterize the genetic complexity of MPN pathophysiology, and 2) to understand and control the progression of MPNs from chronic to life-threatening stages.

We have developed a new mouse model of MPNs by hematopoietic-selective (Mx1-cre) activation of oncogenic Ras (NRas^G12D+/-) and inactivation of Ezh2, the enzymatic subunit of the Polycomb Repressive Complex 2 (PRC2), which mimics concurrent mutations in human MPNs. While NRas^G12D+/- alone led to an indolent myeloproliferative disorder, Ezh2 loss markedly accelerated disease progression to life-threatening stages, resulting in thrombocytosis, neutrophilia, splenomegaly, primary myelofibrosis, leukemic transformation and mortality. Strikingly, while loss of Ezh1, a homolog of Ezh2, had no effect on normal hematopoiesis, concurrent inactivation of Ezh1 and Ezh2 with NRas^G12D+/- completely abolished MPN development and myelofibrosis in vivo . These results establish distinct functions of Ezh1 and Ezh2 in myeloid neoplasms, and identify a selective epigenetic vulnerability for MPNs induced by Ezh2-deficiency. Follow-up studies of Ezh1/2-mediated chromatin (by ChIP-seq) and transcriptomic (by RNA-seq) changes in normal HSCs and leukemia-initiating cells (LICs) revealed non-overlapping gene targets and cellular pathways required for the distinct cellular phenotypes.

Reprogrammed cellular metabolism underlies the progression of various human cancers including leukemia. We noted by RNA-seq transcriptomic profiling that Bcat1, the first and rate-limiting enzyme catalyzing the transamination of branched-chain amino acids (BCAAs), was aberrantly activated in Ezh2-deficient leukemia-initiating cells (LICs). By ChIP-seq analysis, we observed that the Bcat1 promoter is highly enriched with tri-methylated H3-Lys27 (H3K27me3), suggesting that it is epigenetically silenced by Ezh2-PRC2 in normal HSCs. Furthermore, genetic or pharmacologic inhibition of BCAT1 activity selectively abrogated the clonogenic activity of Ezh2-deficient LICs in vitro and the leukemogenic potential in transplanted recipients in vivo, highlighting a metabolic dependency for Ezh2-mutant myeloid neoplasms. Follow-up targeted metabolomic analysis of Ezh2-deficiency-induced metabolic changes established the critical role of Bcat1 and the associated BCAA metabolism in the progression of myeloid neoplasms and leukemia.

Collectively, our results not only elucidate the mechanisms by which Ezh1 is specifically required for Ezh2-deficient LICs, but also validate an important epigenetic vulnerability for MPN patients carrying loss-of-function Ezh2 mutations. Moreover, in-depth analysis of BCAA metabolism in myeloid neoplasms identifies a new metabolic dependency for Ezh2-deficient LICs. Because normal hematopoietic stem cells are unaffected by loss of Ezh1 or Bcat1, the selective vulnerabilities raise the possibility of leveraging Ezh1 or BCAA metabolism as targeted therapies to specifically eradicate Ezh2-mutant LICs. Hence, our study promises to provide critical insights into developing new and generalizable therapeutic strategies to selectively eliminate leukemia-initiating cells harboring alterations of epigenetic pathways.