Physiologic Hypoxia Promotes Maintenance of CML Stem Cells Despite Effective BCR-ABL Inhibition

Chronic myelogenous leukemia (CML) is a disease of hematologic stem cells caused by the BCR-ABL kinase. The use of the ABL kinase inhibitor, imatinib (IM), has dramatically improved the control of CML, but does not cure it. Current evidence suggests that CML stem cells are able to survive IM treatment, and initiate disease recurrence when IM therapy is stopped. The eradication of residual CML stem cells will depend on the identification of mechanisms that support their persistence. The bone marrow microenvironment supports the growth and renewal of hematopoietic stem cells (HSC), and possibly, CML stem cells. Recent studies reveal that physiologic hypoxia (0.1–10%) is a critical component of the microenvironment. We hypothesize that hypoxia maintains CML stem cell function, and that it can do so despite effective BCR-ABL kinase inhibition. To test this hypothesis, we characterized the effect of hypoxia on immunophenotypically-defined CML progenitor and stem cells in the presence and absence of IM. To examine the effect of hypoxia on CML progenitors, we first performed colony forming assays (CFA) on primary chronic phase (CP) CML cells treated with IM (0, 0.25, 1 and 5 μM) under hypoxic (0.5 O₂) or normoxic (21% O₂) conditions for 96 hours. We found that hypoxia increased the number of colonies by 1.6- and 3.2- fold compared to 21% O₂ at 0 and 5μM IM respectively, suggesting that hypoxia favors the maintenance of CML progenitors. To explore the relationship between cell proliferation and hypoxia, we employed CFSE staining to track cell divisions. We found that while both CD34⁺CD38⁺ (committed progenitors) and CD34⁺CD38⁻ (stem cells) CP CML cells proliferated more slowly under hypoxia alone (∼20% decrease in the proliferation index), hypoxia did not further enhance the anti-proliferative effects of IM in both populations. Together, these observations suggest that hypoxia enhances the clonogenic activity of CML progenitors, and that it is able to do so despite the anti-proliferative effects of IM. To rule out the possibility that hypoxia impaired the ability of IM to inhibit BCR-ABL kinase, we also examined the level of phospho-CrkL by intracellular flow cytometry in the CD34⁺CD38⁺ and CD34⁺CD38⁻ CML populations. We found that IM reduced the level of pCrkL in both CD34⁺CD38⁺ and CD34⁺CD38⁻ CML cells, and that hypoxia had no effect on the degree of pCrkL reduction. In line with this result, we also found that hypoxia had no effect on the degree of apoptosis induced by IM. These results indicated that hypoxia increased the clonogenic activity of CML progenitors, and that this occurred in a BCR-ABL-independent manner. These data encouraged us to determine if hypoxia had similar effects on the CML stem cell, which we assayed by performing limiting dilution LTCIC assays, and by assessing the ability of hypoxia-exposed CML cells to engraft immunodeficient NSG mice. In both assays, CML cells incubated under hypoxia showed increases in stem cell capacity (Figure 1). Taken together, our results indicate that physiologic hypoxia maintains the ability of CML cells to act as leukemia stem cells (LSC), and that this can occur in a BCR-ABL-independent manner. In this respect, the functional effects of hypoxia on CML stem cells mirror those on normal HSCs (Danet et al. JCI, 2003). To provide support for this notion, we performed gene expression analysis using Affymetrix U133 Plus 2.0 chips on CD34⁺ cells obtained from normal individuals, as well as CML patients in CP and blast crisis (BC). Preliminary analyses of the microarray data suggest that the hypoxic response in normal and CD34⁺ CP cells was remarkably similar, in contrast to the CD34⁺ BC response. Together, our data indicate that: 1. Physiologic hypoxia enhances the capacity of CP CML progenitors to function as LSCs; 2. Hypoxia enhances CP LSC function independently of BCR-ABL; 3. CP CD34⁺ CML cells have a similar gene expression response to hypoxia compared to normal CD34⁺ cells. Our findings suggest that therapies targeting BCR-ABL will not eliminate CP CML stem cells, particularly in the context of a hypoxic microenvironment.