Bmi1 Is Essential for Leukemic Reprogramming of Myeloid Progenitor Cells

Abstract 1423 Poster Board I-446 The polycomb group (PcG) protein Bmi1 plays an essential role in the maintenance of self-renewing hematopoietic stem cells (HSCs). Derepressed p16Ink4a and p19Arf are tightly associated with a loss of self-renewing capacity of HSCs in Bmi1-deficient mice. Deletion of both Ink4a and Arf genes substantially restores the self-renewal capacity of Bmi1−/− HSCs. Thus, Bmi1 maintains HSCs by acting as a critical failsafe against the p16Ink4a- and p19Arf-dependent senescence pathway. Meanwhile, Bmi1 was originally identified as a collaborating oncogene in the induction of lymphoma and was subsequently reported to be overexpressed in various human cancers including leukemia. Recent studies have demonstrated that PcG proteins bind to multiple regions of the genome and regulate a bunch of target genes. Therefore, we asked whether Bmi1 is essential for leukemic stem cells (LSCs) and tried to identify critical target genes for Bmi1 other than Ink4a and Arf in leukemia. We expressed the MLL-AF9 leukemic fusion gene in purified Lin−Sca-1−c-Kit+CD34+FcγRII/ IIIhi granulocyte/macrophage progenitors (GMPs) from wild-type, Bmi1−/−, Ink4a-Arf−/−, and Bmi1−/−Ink4a-Arf−/− mice and performed in vitro myeloid progenitor replating assay. GMPs from 4 different genetic backgrounds were all immortalized in vitro, although Bmi1-deficient cells showed a slightly decreased replating efficiency. We then infused the immortalized cells into lethally irradiated recipient mice. Mice infused with wild-type and Ink4a-Arf−/− cells developed acute myelogenous leukemia (AML) at 30 to 60 days after infusion. Mice infused with Bmi1−/− cells did not develop leukemia at all. While a significant portion of mice infused with Bmi1−/−Ink4a-Arf−/− cells developed AML, although they took much longer time compared to those mice infused with wild-type and Ink4a-Arf−/− cells. These results indicate that as in HSCs, the Ink4a /Arf locus is one of the major targets for Bmi1 in leukemogenesis. In order to find unknown targets of Bmi1 in LSCs, we compared gene expression profiles of purified c-KithiFcRγII/IIIhiCD34+ cells from Ink4a-Arf−/− and Bmi1−/−Ink4a-Arf−/− immortalized cells. We found that the loss of Bmi1 did not affect the induction of MLL-AF9 target gene expression. By contrast, a number of genes were derepressed in the absence of Bmi1. Among these, Tbx15, a transcriptional co-repressor gene, appeared to be regulated by Bmi1 and a potential tumor suppressor gene in the development of leukemia. Of interest, the majority of derepressed target genes in transformed Bmi1−/−Ink4a-Arf−/− cells, including Tbx15, remained unchanged by re-expression of Bmi1. Correspondingly, re-introduction of Bmi1 to transformed Bmi1−/−Ink4a-Arf−/− cells failed to rescue their compromised leukemogenic activity in vivo. Our findings suggest that Bmi1 is required for faithful epigenetic reprogramming of myeloid progenitors into LSCs by leukemic fusions and contributes to establish LSC-specific transcriptional profiles to confer full leukemogenic activity on LSCs. Disclosures: No relevant conflicts of interest to declare.