Functional Characterization of Epigenetic Modifiers of Demethylating Therapy in AML

Acute myeloid leukemia (AML) is the most common type of acute leukemia in adults, and remains a difficult to treat disease, especially in elderly patients. AML arises in hematopoietic stem and progenitor cells and frequently carries alterations in genes involved in epigenetic regulation. Furthermore, azacitidine (AZA) and decitabine (DAC), two nucleoside analogs inhibiting DNA methylation (DNMTi), have shown clinical efficacy in the therapy of myelodysplastic syndrome (MDS) and AML, highlighting the importance of epigenetic regulation in AML. Up to now, the mechanism of action of AZA and DAC in AML has not been fully elucidated, and reliable biomarkers of therapy response to these drugs do not yet exist. To identify epigenetic response modifiers towards AZA and DAC treatment, we conducted a pooled shRNA-based screen in three human leukemia cell lines using an inducible custom shRNA library targeting more than 660 different genes involved in epigenetic pathways. After selection of retrovirally infected cells, shRNA expression was induced and a fraction of cells were treated with either AZA, DAC or left untreated. 10 days after shRNA induction, genomic DNA was extracted and deep-sequenced to identify genes conferring a selective advantage/disadvantage under these conditions. By comparing shRNA representation changes between the differentially treated groups, we identified multiple candidate genes sensitizing or protecting leukemic cells to/from AZA and DAC treatment. Notably, the DNMT1 gene itself expectedly emerged as an important sensitizer to DNMTi, as did other CXXC domain-containing genes binding to (un-)methylated CpG domains. Another prominent set of genes in the pooled analysis belonged to the DNA damage pathway, including genes such as PARP1 and BRCA1. The identified candidate genes were validated and confirmed in more than 80% when tested in single shRNAs competition assays. Their impact on DNA methylation and gene expression is currently being assessed, with ongoing experiments focusing on the mechanistic role of some of the validated target, as well as the clinical potential for combinatorial therapies. Thus, using an unbiased functional genetic approach, we identified multiple genes including CpG binding proteins and members of the DNA damage pathway as important modulators of DNMTi response. Our results will increase our understanding of the molecular mechanisms driving DNMTi efficacy, and may help to identify new biomarkers and novel targets f