Hypoxia-Associated Effects on Reactive Oxygen Species Generation by Human Acute Myeloid Leukemia Cells

Abstract 4998

Reactive oxygen species (ROS) reflective of oxidative stress response play conflicting roles in cancer biology and therapy. Elevated ROS levels have been implicated in carcinogenesis via DNA damaging effects and activation of pro-survival pathways. In acute myeloid leukemia (AML) patient samples, high ROS levels have been associated with increased risk of relapse and poorer survival following conventional chemotherapy. However, several agents with known anti-leukemic activity have also been shown to mediate anti-tumor effects by inducing oxidative stress in association with cancer cell apoptosis and death.

Recent evidence has suggested that the marrow microenvironment harboring AML cells in vivo is characterized by intrinsic hypoxia. Here we asked if oxidative stress responses by AML cells were potentially altered under intrinsically hypoxic microenvironment conditions as well as following treatment with cytarabine and sorafenib. ROS generation was assessed via fluorescent flow cytometric measurements of CM-H2DCFDA in two human AML cell lines (HL60-VCR, HEL) cultured under normoxic (O₂ 21%) vs. hypoxic (O₂ 1%) conditions for up to 72 hours. Our results revealed higher levels of ROS production in AML cells (HEL, HL60-VCR) cultured under progressively longer periods of hypoxia up to 72 hours. To determine whether this effect was mediated by hypoxia inducible factor-1α (HIF-1α), a transcription factor involved in the hypoxic responses of both normal and cancer cells, ROS generation was measured in normoxic AML cells following treatment with the prolyl hydroxylase inhibitor DMOG which prevents HIF-1α degradation and results in HIF-1alpha protein overexpression. DMOG treatment (0.1-0.3 nM) of HEL and HL60 cells failed to alter ROS levels in patterns similar to what was observed under hypoxia, indicating that hypoxia-induced ROS production likely did not occur primarily via a HIF-1a dependent mechanism. Hypoxia-induced ROS production in AML cells also did not appear dependent on RAC1, a G-protein involved in the oxidative responses of normoxic AML cells and other normal hematopoietic cells. Next we examined the effects of cytarabine treatment on ROS generation by AML cells under differing oxygen conditions. Although short-term cytarabine treatment (up to 48 hours) was associated with mild oxidative stress in AML cells, we noted that cytarabine-treated AML cells exposed to 72 hours of hypoxia continued to exhibit ROS levels similar to those observed under normoxia. We then examined the effects of sorafenib, a receptor tyrosine kinase inhibitor previously reported to induce apoptosis of cancer cells via mitochondria-dependent oxidative stress responses, on AML cells. As compared with vehicle or cytarabine, sorafenib treatment was associated with markedly enhanced ROS production under normoxia; however, under chronic hypoxia, ROS generation by sorafenib was significantly abrogated to below baseline normoxia levels after 48–72 hours.

These results suggest that a hypoxic marrow microenvironment may promote AML growth and therapy resistance in vivo via mediation of specific oxidative stress responses. Our data show that duration of chronic hypoxia progressively increased baseline ROS generation in AML cells and could explain the high levels of ROS found at relapsed AML samples. Moreover, our finding that attenuation of cytarabine/sorafenib-induced ROS generation occurred under the same prolonged hypoxic conditions where decreased chemotherapy-mediated cell death was noted (Hsu et al, ASH abstract 2010) implies a potential association between reduction in oxidative stress and therapeutic responses. As ROS generation under hypoxic conditions did not appear to be primarily mediated by HIF-1a or RAC1, further studies exploring the underlying pathways responsible for oxidative stress responses under chronic hypoxia in AML cells and primary patient samples are warranted.