The Hypomethylating Agent 5-Azacytidine Can Target Leukaemic Stem and Progenitor Cells In Acute Myeloid Leukemia (AML) by Inducing DNA Damage and Stress Response Pathways

Abstract 292

Acute Myeloid Leukemia (AML) is a heterogeneous disease that arises from malignant leukaemic stem and progenitor cells (LSPC). LSPC are relatively resistant to current chemotherapy and considered to contribute to disease progression and relapse in patients due to their ability to re-initiate long-term leukaemic cell growth following initial treatment. We have previously shown that CD34⁺CD38⁻CD123⁺ LSPC are less sensitive to cytarabine and to the FLT3 inhibitor AG1296 whilst exhibiting differential sensitivity to Mylotarg (Gemtuzumab Ozogamicin) when grown under niche support, compared to LSPC with no niche support. As the hypomethylating agent and nucleoside analog 5-Azacytidine (Aza) is showing promise clinically in AML and MDS patients with effects, including induction of ATM, seen as early as 5 days post chemotherapy, we assessed this agent in our model. LSPC from AML samples treated for 72hrs with Aza (1microM) were found to be relatively cytotoxic, resulting in 33% cell kill (n=15) compared to 15% AML bulk cell kill (n=19; p=0.031). We used a protein profiler assay to depict the expression of 46 known human phosphokinase proteins when treated with Aza using the CD34⁺CD38⁻ TF1A cell line. Two phosphokinase proteins were highly activated upon Aza treatment and these were the cell cycle checkpoint regulator and putative tumor suppressor Chk2 and the stress-induced p38 mitogen-activated protein kinase (p38MAPK). Drug-induced DNA damaged cells are known to stall the cell cycle to allow DNA repair by activating the traditional ATM/Chk2/ATR/Chk1 pathway and the more recently described ATM/ATR/p38MAPK/MK2 checkpoint pathway, known to be essential in p53 deficient cells. Thus, we next aimed to examine the DNA damage and stress pathways activated by Aza treatment. Using several AML cell lines, including the CD34⁺CD38⁻ KG1A and TF1A cells, we demonstrate that Aza treatment eliminates S-phase cells while inducing massive DNA damage response (increase in gH2AX), leading to the activation of the ATM/Chk2 and Chk1 and thus triggering G2/M cell cycle arrest. To date 3 primary AML samples have been treated in vitro with Aza and downstream effects analysed: only 1/3 was sensitive to Aza and this was the only sample that showed an increase in ATM and Chk2 in response to the Aza treatment. As we wanted to investigate whether activation of the stress response pathway encourages any additional cell survival, we pre-treated the same cell lines with the p38MAPK inhibitor SB203580 for 30minutes prior to Aza treatment. Our data shows that there is enhanced cell kill if the p38MAPK pathway is inhibited first and then Aza treated later. We are currently in the process of assessing induction of downstream effector proteins involved in the p38MAPK pathways after Aza treatment. In conclusion, the success of Aza treatment in AML is maybe due to the specific targeting of the relapse-causing LSPCs that is achieved by the activation of DNA damage and stress response pathways leading to G2/M cell cycle arrest. We are currently assessing the efficacy of Aza in in vivo animal models and investigating any dysregulated methylation patterns in Chk2 and p38MAPK in AML cells.