Abstract 2986

Acute myelogenous leukemia (AML) is a fatal disease in which the majority of patients relapse and die even after attaining initial complete remission. AML is thought to be initiated and maintained by chemoresistant leukemia stem cells (AML-SCs). Therefore, identifying therapies that can eliminate AML-SCs is a priority.

Iron is crucial to normal cell metabolism and plays a role in multiple cellular activities, including promotion of processes required for maintenance of malignancy, such oxidative phosphorylation, nucleotide synthesis, and Wnt signaling. An important regulator of intracellular iron is ferroportin, an iron exporter. Using a publicly available gene expression dataset for AML patients (GEO accession #GSE6891), we found that low levels of ferroportin correlate with poor outcomes (p = 0.018). We investigated the levels of ferroportin in AML-SCs and discovered that ferroportin levels are significantly lower in AML-SCs than in their normal counterparts (p=0.008). Based on these findings, we hypothesized that aberrant iron metabolism may be an important feature of LSC biology that could be targeted by a novel therapeutic agent. To this end, we investigated the activity of known natural products with iron chelation activity in AML-SCs, specifically focusing on proanthocyanidins found in cranberry extracts. Many of the reported health benefits of cranberries, including their antimicrobial functions, are associated with a unique class of proanthocyanidins referred to as A–type PACs (A-PACs). We tested the effects of a commercially available cranberry extract (Cysticran 40; CYS) and A-PACs in 15 primary AML and 5 normal CD34+ cord blood specimens and found potent and specific anti-LSC activity.

Primary AML samples were shown to be highly sensitive to CYS, with a mean LD50 of 180.6 μg/ml (110.2 – 251.1 μg/ml, 95% CI; n=9). Purified PACs demonstrated even greater potency (mean LD50= 82.51 μg/mL; 57.07–107.9, 95% CI; n=11). The sensitivity to PACs and CYS was also observed in phenotypically described progenitor and stem cell populations from the AML samples. Sensitivity to CYS and A-PACs was not confined to AML with specific cytogenetic abnormalities or known mutations, suggesting potency across AML subtypes. Importantly, we did not observe any overt effects on purified CD34+ cells from healthy cord blood samples. Functional stem cell assays showed ablation of AML-SCs with A-PAC treatment. Specifically, primary AML samples treated with 62.5μg/ml demonstrated more than 75% decrease in colony forming activity relative to vehicle control (n=5). In contrast, there was less than 2 fold decrease in colony formation in CD34+ CB cells treated with 125μg/ml of PACs (n=4). Xenotransplant assays showed significantly decreased human AML engraftment after treatment with 62.5μg/ml A-PAC (90.6% decreased engraftment, n=3, p<0.001), while normal CD34+ cells retained engraftment capability in immunodeficient mice (n=4).

We observed that treatment with CYS and PACs resulted in caspase-3 activation, evaluated by immunoblots and flow cytometry. Furthermore, pre-treatment with antioxidants or holo-transferrin partially protected AML cells from A-PAC induced cell death (p<0.01). In addition, A-PAC treatment induced changes in cellular iron metabolism and increased ROS levels. Interestingly, gene expression analysis revealed that A-PACs upregulated chemokine and NF-kB pathways (p= 1.2 × 10−9), which is uncharacteristic of anti-LSC compounds discovered to date and suggests a novel mode of AML-SC ablation that bypasses NF-kB signaling to achieve AML-SC ablation.

Together, our results suggest that cranberry A-PACs represent a novel class of compounds with therapeutic potential to ablate leukemia stem and progenitor cells, with minimal effects on normal hematopoietic stem cells.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.