• Hypoxia mediates TKI resistance.• Hypoxia enhances CML stem cell maintenance.C-abl oncogene 1, nonreceptor tyrosine kinase (ABL1) kinase inhibitors such as imatinib mesylate (imatinib) are effective in managing chronic myeloid leukemia (CML) but incapable of eliminating leukemia stem cells (LSCs), suggesting that kinase-independent pathways support LSC survival. Given that the bone marrow (BM) hypoxic microenvironment supports hematopoietic stem cells, we investigated whether hypoxia similarly contributes to LSC persistence. Importantly, we found that although breakpoint cluster region (BCR)-ABL1 kinase remained effectively inhibited by imatinib under hypoxia, apoptosis became partially suppressed. Furthermore, hypoxia enhanced the clonogenicity of CML cells, as well as their efficiency in repopulating immunodeficient mice, both in the presence and absence of imatinib. Hypoxia-inducible factor 1 a (HIF1-a), which is the master regulator of the hypoxia transcriptional response, is expressed in the BM specimens of CML individuals. In vitro, HIF1-a is stabilized during hypoxia, and its expression and transcriptional activity can be partially attenuated by concurrent imatinib treatment. Expression analysis demonstrates at the whole-transcriptome level that hypoxia and imatinib regulate distinct subsets of genes. Functionally, knockdown of HIF1-a abolished the enhanced clonogenicity during hypoxia. Taken together, our results suggest that in the hypoxic microenvironment, HIF1-a signaling supports LSC persistence independent of BCR-ABL1 kinase activity. Thus, targeting HIF1-a and its pathway components may be therapeutically important for the complete eradication of LSCs. (Blood. 2014;123(21):3316-3326)

Cancer cells, including in chronic myeloid leukemia (CML), depend on the hypoxic response to persist in hosts and evade therapy. Accordingly, there is significant interest in drugging cancer-specific hypoxic responses. However, a major challenge in leukemia is identifying differential and druggable hypoxic responses between leukemic and normal cells. Previously, we found that arginase 2 (ARG2), an enzyme of the urea cycle, is overexpressed in CML but not normal progenitors. ARG2 is a target of the hypoxia inducible factors (HIF1−α and HIF2−α), and is required for the generation of polyamines which are required for cell growth. We therefore explored if the clinically-tested arginase inhibitor Nω−hydroxy−nor−arginine (nor−NOHA) would be effective against leukemic cells under hypoxic conditions. Remarkably, nor−NOHA effectively induced apoptosis in ARG2-expressing cells under hypoxia but not normoxia. Co-treatment with nor−NOHA overcame hypoxia-mediated resistance towards BCR−ABL1 kinase inhibitors. While nor−NOHA itself is promising in targeting the leukemia hypoxic response, we unexpectedly found that its anti-leukemic activity was independent of ARG2 inhibition. Genetic ablation of ARG2 using CRISPR/Cas9 had no effect on the viability of leukemic cells and their sensitivity towards nor−NOHA. This discrepancy was further evidenced by the distinct effects of ARG2 knockouts and nor−NOHA on cellular respiration. In conclusion, we show that nor−NOHA has significant but off-target anti-leukemic activity among ARG2-expressing hypoxic cells. Since nor−NOHA has been employed in clinical trials, and is widely used in studies on endothelial dysfunction, immunosuppression and metabolism, the diverse biological effects of nor−NOHA must be cautiously evaluated before attributing its activity to ARG inhibition.