HIF-1 Prevents Hematopoietic Cells from Cell Damage by Overproduction of Mitochondrial ROS after Cytokine Stimulation through Induction of PDK-1

Reactive oxygen species (ROS) play numerous functions in both physiological and pathological hematopoiesis. Although ROS are generated in multiple cellular compartments and by multiple enzymes, the majority is derived from mitochondria coupled with glucose metabolism. To detoxify the cytotoxicity of ROS, cells have several ROS scavenging mechanisms. In addition, recent studies have revealed that hypoxia inducible factor-1 (HIF-1) contributes to the reduction of mitochondrial ROS production through the induction of pyruvate dehydrogenase kinase-1 (PDK-1) in response to hypoxia. PDK-1 phosphorylates and inactivates pyruvate dehydrogenase (PDH) E1α, which converts pyruvate to acetylcoenzyme A. Thus, the induction of PDK-1 by HIF-1 blocks the entry of pyruvate into the tricarboxylic acid (TCA) cycle and subsequently suppresses mitochondrial ROS production under hypoxic conditions. In addition to hypoxia, mitochondrial ROS production is also enhanced by cytokine stimulation. Previously, we showed that thrombopoietin (TPO) induced the generation of mitochondrial ROS and that mitochondrial ROS induced by TPO promoted the elevation of HIF-1α, a subunit of HIF-1, both in UT-7/TPO cells and in primary mouse progenitor cells. Based on these observations, we speculated that TPO-activated HIF-1 works as a feedback mechanism to block overproduction of ROS, which may be harmful to hematopoietic cells, by controlling PDK-1 expression. To investigate this notion, we first analyzed ROS production kinetics after TPO stimulation using UT-7/TPO cells. ROS production was gradually increased, peaking at 48 hr, and then decreasing by 72 hr after TPO stimulation. PDK-1 expression increased 48 hr after TPO treatment before ROS dropped. Phosphorylation of PDH-E1α was enhanced by TPO in the same fashion. To confirm that PDK-1 induction by TPO contributes to reduction of ROS, we treated UT-7/TPO cells with a PDK-1 inhibitor, dichloroacetate (DCA). As expected, DCA blocked the phosphorylation of PDH-E1α and induced sustained ROS production. Furthermore, DCA treatment resulted in increased apoptotic cell ratio after TPO stimulation. These results support our hypothesis that PDK-1 works to prevent overproduction of ROS after TPO stimulation and also suggest that sustained production of ROS after cytokine stimulation might be toxic to hematopoietic cells. Next, we analyzed whether HIF-1 is required for this process. Echinomycin, a specific inhibitor of HIF-1, blocked PDK-1 elevation by TPO in a dose-dependent manner. For further study, we established a stable HIF-1α knockdown system in UT-7/TPO cells using siRNA. TPO failed to induce PDK-1 expression in these clones. Concomitantly, ROS levels in HIF-1α knockdown cells remained high, compared to parental cells stimulated with TPO for 72 hr. We also found sustained activation of p38 MAPK and JNK in HIF-1α knockdown clones after TPO stimulation. Taken together, our observations suggest that TPO-induced activation of HIF-1 and subsequent induction of PDK-1 is an important mechanism to prevent overproduction of mitochondrial ROS, a secondary product of glucose metabolism. This feedback mechanism may be critical for protecting hematopoietic cells from DNA damage by ROS production after cytokine stimulation.