Wild-Type and Mutant C-Kit Activation of Stat3α Contribute to Leukemogenesis through Distinct Effects on Myeloid Cell Proliferation and Resistance to Apoptosis

Expression of a constitutively active mutant of the receptor tyrosine kinase c-Kit in patients with acute myeloid leukemia (AML) is associated with poor survival. Most mutations occur in the activation loop, at Asp 816 or Asn 822, resulting in ligand-independent tyrosine phosphorylation and resistance to many of the tyrosine kinase inhibitors currently in use. Signal transducer and activator of transcription 3 (Stat3) also is upregulated and/or constitutively activated in many malignancies, including AML. Stat3 is activated by both wild-type c-Kit, which is expressed on most AML blasts, and by mutant c-Kit. Like mutant c-Kit, increased Stat3 activity has been associated with poor outcome in patients with AML. Stat3 exists in two isoforms – full length Stat3α (p92) and C-terminally truncated Stat3β (p83) – and the contribution of each isoform to c-Kitmediated leukemogenesis is not known. To begin to address this issue, we generated IL-3-dependent 32D murine myeloid cells that inducibly overexpress Stat3α when treated with dexamethasone and doxycycline (induced: Stat3α^HI; uninduced: Stat3α^LO). We further transfected these cells with wild-type c-Kit (c-Kit^WT), c-Kit^D816H or c-Kit^N822K. First, we established by immunoblotting that c-Kit and Stat3α were tyrosine-phosphorylated only after stem cell factor (SCF) stimulation in cells expressing c-Kit^WT, but were constitutively phosphorylated in c-Kit^D816H and c-Kit^N822K cells. To test the effects of increased Stat3α on survival and proliferation, we withdrew cells from IL-3 and stimulated with SCF (100 ng/ml) for 3 days; control cells remained unstimulated. We measured the proportion of cells undergoing apoptosis at day 3 by FACS using annexin V-PE labeling. In cells expressing c-Kit^WT, apoptosis was significantly reduced in SCF-stimulated Stat3α^HI cells compared to SCF-stimulated Stat3α^LO cells (8.4 ± 5.0% v. 40.8 ± 19.9%; p<0.05). In cells expressing c-Kit^D816H or c-Kit^N822K, apoptosis was significantly reduced in Stat3α^HI cells compared to Stat3α^LO cells (c-Kit^D816H: 30.2 ± 10.2% v. 66.7 ± 8.1%, p<0.05; c-Kit^N822K: 25.1 ± 17.4% v. 61.4 ± 19.8%, p<0.05). SCF treatment had no detectable effect on apoptosis in cells expressing mutant c-Kit. Next, we measured the rate of proliferation using the membrane tracking dye PKH26. Fluorescence histograms, with gating on live cells, were modeled using ModFit LT, and the proliferation index (PI; total number of cells/calculated number of cells in parental generation) at day 3 was calculated. Cells expressing c-Kit^WT did not proliferate well in the absence of SCF-stimulation, with PIs 1.6 ± 0.1 and 2.1 ± 0.3 for Stat3α^LO and Stat3α^HI, respectively. The PI for SCF-stimulated Stat3α^HI cells was significantly higher than the PI for SCF-stimulated Stat3α^LO cells (5.4 ± 0.7 v. 3.7 ± 1.1, p<0.05). Surprisingly, cells expressing mutant c-Kit showed only low levels of proliferation, regardless of Stat3α level or SCF stimulation. The PI was 2.0 ± 0.5 in unstimulated Stat3α^HI c-Kit^D816H cells and 2.4 ± 0.3 in Stat3α^HI c-Kit^N822K cells. Thus, both wild-type and mutant c-Kit activate Stat3α, which results in marked resistance to apoptosis. Wild-type c-Kit also activates a distinct signaling pathway that drives proliferation, and this process is augmented by Stat3α activation. These findings suggest that the critical role for Stat3α in myeloid leukemogenesis is apoptosis prevention and that targeting Stat3α is likely to be an effective therapeutic approach for leukemias with wild-type or mutant c-Kit.