Molecular Mechanism Regulating Foxo In Leukemia Initiating Cells of Chronic Myeloid Leukemia.

Abstract 3391

Chronic myeloid leukemia (CML) is caused by a defined genetic abnormality that generates BCR-ABL, a constitutively active tyrosine kinase. Although the development of imatinib, a small molecule inhibitor of ABL, represents a breakthrough in the treatment of CML, major part of patients treated in chronic phase CML are not off therapy due to resistance or intolerance. Recent studies have suggested that imatinib is a potent inhibitor against differentiated leukemia cells, but does not deplete leukemia-initiating cells (LICs) responsible for recurrence of CML. To date, therapeutics that can eradicate CML LICs, however, have remained under investigation. To overcome these clinical problems, here we studied the molecular mechanisms regulating maintenance of imatinib-resistant CML LICs by forkhead transcription factor Foxo3a.

We first generated a mouse CML model by using retroviral induction of BCR-ABL-ires-GFP gene into mouse immature hematopoietic cells, and the cells were subsequently transplanted into irradiated recipient mice. These experiments showed that CML LICs were highly enriched in c-Kit⁺Lin⁻Sca-1⁺ (KLS⁺) population in BCR-ABL⁺ CML cells. Serial transplantation experiments for CML LICs originated from Foxo3a-deficient mice and littermate wild-type mice indicated that Foxo3a-deficiency reduced lethality of recipient mice at third transplantation. Although recipients that transplanted with wild-type LICs developed CML and acute lymphocytic leukemia (ALL) at third transplantation, we did not observe development of ALL or CML in recipients of Foxo3a deficient LICs after 45 days post-third transplantation, suggesting that the Foxo3a deficient LICs lose their potential to generate malignancies. In addition, a combination of Foxo3a deficiency and imatinib treatment led to efficient depletion of CML in vivo, indicating that Foxo3a plays an essential role for the maintenance of imatinib-resistant CML LICs (Naka et al., Nature 463, 676–680, 2010).

Interestingly, when we examined sub-cellular localization of Foxo3a transcription factor in the CML LICs, we found two CML LIC populations; one population was the cells with nuclear localization of Foxo3a (Foxo3a transcription factor is active) and the other population was the cells with cytoplasmic localization of Foxo3a (Foxo3a is inactive). To understand the molecular mechanisms regulating Foxo3a in CML LICs, we next evaluated the activity of upstream BCR-ABL, PI3K, PDK1, and Akt signaling pathway by fluorescence immunohistochemistry. BCR-ABL activity that was determined by phosphorylation levels of CrkL, a down-stream target of BCR-ABL, was detected in almost all of the CML LICs. However, unexpectedly, phosphorylation levels of Akt in the CML LICs with nuclear localization of Foxo3a appeared to be lower than that in the CML LICs with cytoplasmic localization of Foxo3a, despite it is widely believed that BCR-ABL induces activation of Akt signal. Consistent with Akt phosphorylation status, we detected low levels phosphorylation of PDK1 and PI3K p85, upstream regulators for Akt, in the CML LICs with nuclear localization of Foxo3a. Interestingly, expression levels of the cell proliferation antigen Ki67 were lower in the CML LICs with nuclear Foxo3a than that in the CML LICs with cytoplasmic Foxo3a. These results suggest that Foxo3a responsible for maintenance of imatinib-resistant CML LICs may be regulated by molecular mechanisms that are involved in dormancy in CML LICs.