New Insights into PLZF/Rara Mechanism during APL Onset: EZH2 on the Road

INTRODUCTION Inappropriate recruitment of functional Polycomb-Group proteins (PcG) may trigger epigenetic unbalance at very specific genomic loci that substantially contribute to the pathogenesis of Acute Myeloid Leukemia (AML). This concept was first described in Acute Promyelocytic Leukemia (APL) in which PcG proteins were abnormally addressed due to the expression of X-RARA fusion proteins and were involved in the treatment response of the disease. For instance, in the context of APL with t(11;17)(q23;q21) translocation, the resulted oncogenic fusion protein PLZF/RARA leads to abnormal recruitment of PcG at the promoters of genes involved in acid-trans-retinoic acid (ATRA) response (Boukarabila et al.). As a consequence of that and compared to other APL subtypes (e-g: PML/RARA), APL with PLZF/RARA are insensitive to ATRA. In the recent years, a repertoire of cis-regulatory enhancer elements has been dissected to reveal important insight about leukemia onset and define new subsets of the disease with different treatment responses (Bhagwat et al). As we previously reported that PLZF displayed epigenetic specificity on enhancers (Poplineau et al.) we questioned the role of PLZF/RARA on these regulatory regions during APL onset. METHODS We performed in vivo comparative epigenomic profiling (H3K27ac, H3K4me1, H3K27me3 and H3K4me3 ChIPseq) between normal myeloid progenitors (granulocyte-monocyte progenitors purified from wild-type mice) and PLZF/RARA transformed mouse progenitors (late promyelocytes purified from mice developing APL). To question the role of PcG in APL onset, we used retroviral overexpression of PLZF/RARA and transduced Lineage negative cells from a conditional KO EZH2 mouse model. Transformation was tested by replating assay and cells were characterized by FACS and morphology analyses. We also performed EZH2 pharmacological inhibition using GSK126 and UNC1999 on a human cell line expressing the fusion protein PLZF/RARA. We analyzed the impact of this inhibition on their transcriptomic signature (RNAseq) and their proliferative capacity. RESULTS Upon PLZF/RARA expression and APL progression, specific cis-regulatory enhancer elements were targeted by the H3K27me3 PcG repressive mark. This gain in poised enhancer regions, upon PLZF/RARA expression reflected a reoriented PcG activity, from enhancers regulating developmental processes to those regulating stress and immune responses. To demonstrate the importance of this H3K27me3 switch for APL progression, we investigated the effect of EZH2 loss during PLZF/RARA transformation. Using a conditional KO EZH2 mouse model, we demonstrated that PLZF/RARA required EZH2 activity to efficiently transform progenitors since EZH2 loss promoted differentiation that altered the replating capacity of the PLZF/RARA expressing cells. In addition, EZH2 inhibition by GSK126 revealed some interesting benefits since it sensitized PLZF/RARA transformed progenitors to ATRA treatment. Moreover, inhibition of EZH2 with GSK126 or UNC1999 induced a decrease in the proliferation advantage of a human PLZF/RARA-inducible cell line. This was linked to a change of its transcriptomic signature towards an expression pattern closer to the one observed in the parent cell line. CONCLUSION Taken together, our data showed that PLZF/RARA modifies H3K27me3 profiles at enhancer regions and requires EZH2 activity for APL onset. Finally, our results suggest that EZH2 inhibition could be a new promising therapeutic approach for retinoic-acid resistant APL. Disclosures No relevant conflicts of interest to declare.