Mutated FLT3 defines a promising target for the treatment of acute myeloid leukemia (AML) with specific protein tyrosine kinase (PTK) inhibitors. The clinical efficacy of this approach, however, is limited due to molecular mechanisms that remain to be elucidated. As we demonstrated previously, overexpression of antiapoptotic proteins of the BCL2 family lead to resistance against PTK inhibitors in cell lines with activating FLT3 mutations (Bagrintseva, Blood, 2005). In primary AML samples tested so far in our study, we found a correlation of the expression level of BCL-XL, a known downstream target of FLT3, and the presence of activating FLT3 mutations whereas MCL1 protein, another antiapoptotic member of the BCL2 family, showed very low expression. In contrast, we found a high expression level of BCL2 protein in all AML samples whether or not FLT3 mutations were present and this level remained unchanged after dephosphorylation of mutated FLT3. Thus, we speculated that an overexpression of BCL2 independent from FLT3 activation might at least in part explain the limited clinical efficacy of PTK inhibitors in the treatment of FLT3 positive AML. To test this hypothesis, we stably expressed mock, BCL2 or BCL-XL, respectively, in Ba/F3 cell lines carrying constitutively activated FLT3 with internal tandem duplication. The cells overexpressing BCL2 or BCL-XL, respectively, did not respond to treatment with the FLT3 specific inhibitor SU5614 up to high doses. To overcome the observed resistance, we tested the small molecule inhibitor ABT-737 (kindly provided by Abbott Laboratories) that has been described to efficiently disrupt intracellular BCL2 family interactions by binding to the hydrophobic BH3 groove of these proteins (Oltersdorf, Nature, 2005). Surprisingly, treatment of our generated cell lines with ABT-737 alone did not result in increased levels of apoptotic cell death. This finding is in line with previous reports showing that mono-treatment with ABT-737 does not directly activate proapoptotic proteins, but needs activator BH3-only proteins such as BID or BIM. Co-treatment of the cell lines with SU5614 and ABT-737, however, rendered them again susceptible to the PTK inhibitor in a concentration-dependent manner. SU5614 and ABT-737 showed synergism as confirmed by immunoblotting against cleaved and full-length caspase-3. As a negative control to all our experiments, we used the functionally inactive enantiomer of ABT-737 (ABT control) that caused significant cytotoxicity neither alone nor in combination with SU5614 up to high doses.

To underline the clinical relevance of these findings, a panel of AML patient samples is currently tested for response to ABT-737 alone or in combination with PTK inhibitors. Two AML samples tested so far showed an IC50 of 10 and 25nM (vs. 300 and 1000 nM, respectively, for ABT control) after 24h of mono-treatment with ABT-737, whereas peripheral blood mononuclear cells of a healthy donor showed an IC50 of 80 nM for ABT-737 and 400nM for ABT control. This might be explained by recent findings indicating that native tumor and leukemia cells are addicted to the expression of antiapoptotic proteins and tonically exposed to proapoptotic stimuli (Certo, Cancer Cell, 2006).

Since BCL2 has been reported to be involved in cell cycle regulation by facilitating G0/G1 arrest, we are also going to study the effects of ABT-737 on non-proliferating CD34+ progenitor AML cells that cannot be eliminated by conventional chemotherapy.

Disclosure: No relevant conflicts of interest to declare.

Author notes


Corresponding author