Despite high remission rate after therapy, only 30-40% of acute myeloid leukemia (AML) patients survive 5 years after diagnosis. The main cause of treatment failure is thought to be insufficient eradication of leukemia initiating cells (LIC). Identifying drugs that can efficiently eradicate LICs is therefore imperative. HIF1α is essential for LIC maintenance and targeting HIF1α selectively eliminates LIC. Deferasirox is an iron chelator used to reduce chronic iron overload in patients receiving long-term blood transfusions. Our aim is to study the ability ofdeferasirox to induce apoptosis and to target HIF1α in AML LICs.

CD34+CD38- LIC-like cells, sorted from the KG1a cell line, exhibited increased sensitivity to deferasirox with an IC50 of 1.3μM compared to 8.9μM calculated for the more mature CD34+CD38+ cells. The LIC-like cells, which were more sensitive to deferasirox, were less sensitive to ARA-C compared to the CD34+CD38+ cells. Deferasirox was >2-fold more efficient in inducing apoptosis in the CD34+CD38 cells, compared to the CD34+CD38+ cells (74±7.2% & 32±6.2% apoptosis, respectively). Deferasirox demonstrated a synergistic cytotoxic effect with ARA-C on both the CD34+CD38 and CD34+CD38+ KG1a cell fractions. Similar results were observed with CD34+CD38-CD123+ LICs and CD34+CD38+ CD123+ progenitor cells isolated from AML patients. As shown with the cell line, AML patient LICs were 2-fold more sensitive to deferasirox treatment showing a 62±15% induction of cell death compared to only a 34±9% induction in leukemic progenitor cell death (p<0.01). The increase in cell death was accompanied by an increment of reactive oxygen species (ROS) levels which was more prominent in the LICs compared to the progenitor cells. Furthermore, deferasirox enhanced the cytotoxic effects of ARA-C on both the LSCs and the leukemic progenitor cells. These data indicate that deferasirox is highly cytotoxic to AML cells; however, it is significantly more specific to AML initiating cells. In addition, deferasirox exhibited significant inhibitory effect on the proliferation of both the CD34+CD38 and CD34+CD38+ KG1a cell fractions accompanying by an S-phase arrest. The inhibition of proliferation was also demonstrated using proliferation dye and colony assay on both the LIC-like cells and the LICs isolated from AML patient (p<0.001). Since deferasirox is an NFkB inhibitor which regulates HIF1α levels, and since HIF1α is selectively activated in AML stem cells and is essential for maintenance of these cells, we studied the effect of deferasirox on this pathway. We found that NFkB translocation to the nucleus was reduced by over 80% and as a result the binding of NFkB to the HIF1α promoter was decreased in the LIC-like cells treated with deferasirox. HIF1α was downregulated transcriptionally (~45% reduction) and translationally (~55% reduction) in the LIC-like cells while it was upregulated in the CD34+CD38+KG1a cells following deferasirox treatment. In order to understand which of deferasirox's properties is crucial for its apoptotic effect, we used the iron chelator, deferipron, which does not effect on NFkB or HIF1α and siHIF1α. Each one of these treatments alone did not lead to an apoptotic effect in LIC-like cells or in the CD34+CD38+ kg1a cells. Conversely, the combination of the two lead to an increase in apoptosis in the LIC-like cells suggesting that only a combination of iron chelation and HIF1α reduction is sufficient to result in an apoptotic response. The apoptotic effect wasn't observed in the CD34+CD38+ fraction. These data hint that AML-LICs are selectively targeted by deferasirox.

We describe a novel and unique anti-LIC property of deferasirox, which was originally developed as an iron chelator. We found clinically relevant concentrations of deferasirox to be cytotoxic in vitro to AML progenitor cells but even more potent against LICs. We believe that deferasirox exerts its cytotoxic effect, at least in part, by downregulating HIF1α levels in the LIC population. Pending further characterization, deferasirox can be considered as a potential therapeutic agent for eradicating LICs.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution