Functional screening of primary DNMT3A-mutant AML cells in search for new therapeutic targets

Acute myeloid leukemia (AML) is a malignancy of the hematopoietic system caused by somatic mutations that accumulate in hematopoietic stem and progenitor cells. The cells are thereby transformed into leukemic stem cells (LSCs), which cannot be efficiently eliminated with the standard chemotherapy treatment. Thus, LSCs pose a risk of relapse for AML patients. There- fore, identification and characterization of LSCs is a major challenge in the field of AML research. Through next generation sequencing approaches the mutational spectrum of AML cells has been established and a continuous effort is being made to resolve the order of mutation acquisition and their functional consequences. In the subgroup of AML patients that bear a mutation in Nucleophosmin 1 (NPM1 ), a mutation in DNA-methyltransferase 3 A (DNMT3A) has been often found as a co-occurring event. Evidence suggests that this mutation arises early in leukemogene- sis and marks leukemic progenitors and stem cells. However, the functional consequences of this mutation are far from being understood. In this thesis work, I set out to unravel novel functional dependencies of the DNMT3A-mutant AML cells that can be exploited for therapeutic purposes. To nominate genes that are essential for the survival of primary AML cells, I performed a func- tional RNA interference-mediated drop out screen in 38 DNMT3A- and NPM1-mutant AML patient lines. The patients in this cohort were divided into two groups, based on the treatment outcome: an early relapse (ER) group and a long term remission (LTR) group. To nominate can- didate genes in each group, I have selected 12 screens with the highest data quality and performed a differential bioinformatic analysis. The analysis yielded 7 potential candidates, from which I initially validated three: Glucocorticoid modulatory element binding protein 1 (GMEB1), Mouse double minute 4 (MDM4) – both shared between the ER and the LTR groups – and Thioredoxin domain containing protein 9 (TXNDC9 ), which scored only in the ER group. Additional rounds of validation nominated MDM4 as the strongest candidate. To investigate the role of MDM4 in LSCs, I knocked it down in three patient samples and performed the long-term culture-initiating cell assay. However, the number of progenitor colonies that formed by the end of the assay was not enough for a statistical evaluation, probably due to the low frequency of long-term culture- initiating cells in the samples. Therefore, no conclusion regarding the functional dependency of LSCs on MDM4 could be made. However, a recent study suggested that loss of MDM4 causes cell cycle arrest and induces apoptosis in leukemic cell lines and primary cells, including progenitor populations, confirming the findings of this thesis. Nevertheless, the question about the role of MDM4 in NPM1 -mutant AML cells remains open. The NPM1 involvement in the p14Arf-MDM2- p53 pathway and the deregulation of this pathway caused by the NPM1 mutation indicate that MDM4 might poses special functions in NPM1 -mutant AML. Therefore, it should be investigated if MDM4 is a particularly suitable therapeutic target in AML with NPM1 mutation