Abstract

Studies have shown that human acute myelogenous leukemia (AML) originates from a rare population of leukemic stem cells (LSCs). LSCs are found in nearly all AML subtypes and are sufficient to initiate and maintain leukemic growth in both long-term cultures and NOD/SCID mice. Since conventional chemotherapy drugs typically target actively cycling leukemic blast cells, the quiescent LSC population is likely to be drug resistant. Recently, we have described agents that can selectively kill LSC while sparing normal hematopoietic stem cells (HSC), thus demonstrating that quiescent LSC can be effectively destroyed. In particular, the sesquiterpene lactone parthenolide (PTL) is active as a single agent and can induce robust LSC-specific apoptosis. Therefore, we sought to further elucidate the mechanisms that result in leukemia-specific death. To this end, we have shown that PTL inhibits NF- κB and activates p53 in AML cells. Moreover, we have found that AML cells are highly-oxidized relative to their normal less-oxidized counterparts as determined by dihydrofluorescein diacetate (H2DCF-DA) fluorescence. Although PTL treatment was found to increase the basal oxidative load of both normal and leukemic cells, these increases were ultimately associated with only leukemia-specific death. High levels of reactive oxygen species are associated with genotoxic stress. Therefore, we hypothesized that PTL-mediated increases of the oxidative load in AML cells would be sufficient to produce genotoxic stress in AML cells, but not normal cells. We thus examined pathways involved in sensing and responding to oxidative and genotoxic stress. These studies demonstrate that upon PTL treatment, AML cells phosphorylate p53 on serine 15, and increase transcription of several downstream p53 pro-apoptotic genes, including GADD45, Bax, and p21. In addition, we detected phosphorylation of ATM on serine 1981, γ-H2AX, and phosphorylation of Chk2 on threonine 68. Importantly, these events are not observed in normal hematopoietic cells treated with PTL. These findings suggest a DNA damage/repair response to PTL in AML cells, and that normal cells appear to be resistant to this mechanism. Collectively, these data indicate that oxidative stress and DNA damage pathways may be a central component of PTL-induced apoptosis. We propose a model whereby AML cells exist in a physiological state in which PTL is able to preferentially invoke genotoxic stress in AML. Based on empirical evidence, we suggest that AML LSC also exist in such a state, and that PTL mediated inhibition of NF- κB and induction of stress leads to selective apoptosis of malignant cells, sparing normal cell populations.

Author notes

Corresponding author