FLT3-ITD+ AML Blast, Progenitor and Stem Cell Populations Demonstrate Higher Sensitivity to the Hsp90 Inhibitor PU-H71,

Abstract 3500

Hsp90, one of the best-characterized molecular chaperones, plays indispensable roles in folding and assembly, intracellular transport, stabilization, and degradation of proteins, and therefore, facilitating cell signaling. Hsp90 is also involved in tumorigenesis by stabilizing oncogenic client proteins. Thus, Hsp90 inhibition has been considered a promising therapeutic strategy for different types of cancer including leukemia. PU-H71 is a novel HSP90 inhibitor with high specificity for oncogenic Hsp90. We investigated the effect of PU-H71 in acute myelogenous leukemia (AML), in particularly, AML stem cells (AML-SCs) that are known to give rise to AML blasts, are refractory to conventional therapies, and thus likely to account for AML relapses.

The effect of PU-H71 was evaluated using a panel of 12 leukemia cell lines. Among the 12 leukemia cell lines tested, MOLM-13 and MV4-11 cells were the most sensitive (LD50s 253 nM and 120 nM respectively). Both MV4-11 and MOLM-13 carry FLT3-ITD mutation (occurring in ∼40% AML cases) and MLL translocations (occurring in ∼20% AML cases). Both MLL and FLT3 have been reported as client proteins of Hsp90. However, other leukemia cell lines with MLL rearrangements such as THP-1 and a MLL-ENL cell line derived from MLL-ENL transformed human CD34+ cord blood cells exhibited resistance to PU-H71 treatment (LD50 > 2 μM). The data suggested that FLT3-ITD+ AML samples may display higher sensitivity to Hsp90 inhibition.

To confirm the higher sensitivity of FLT3-ITD+ AML cells to Hsp90 targeted therapy, 15 primary AML patient samples (8 FLT3-ITD mutants and 7 wild type FLT3) were treated with increasing concentrations of PU-H71. Cell viability on different cell populations was evaluated using multiparameter flow cytometry at 48 hours after treatment with PU-H71. The average LD50 of PU-H71 in FLT3-ITD+ AML cells was 492 nM (95% CI, 127.636 – 856.364). In contrast, the average LD50 in AML samples with wild type FLT3 was 2.795 μM (95% CI, 1.058 – 4.532). The near 6-fold difference between LD50s for PU-H71 was significant (p=0.0068). Importantly PU-H71 also killed FLT3-ITD+ AML stem and progenitor cells more effectively. Furthermore, PU-H71 treatment decreased the ability to form colonies in FLT3-ITD+ AML specimens more effectively than FLT3 WT AMLs (97.6% and 79.3% decrease relative to control respectively; N=3; P=0.0236). Importantly, PU-H71 had minor toxicity to normal blood mononuclear cells and normal cord blood hematopoietic stem cells.

FLT3-ITD+ cell lines and primary AML cells treated with 0.5 μM PU-H71 showed a substantial decrease of phosphorylated forms of Erk1/2, JNK, AKT, p70RSK, NF-κB(p65) and Stat5, which was observed within 4 hours post PU-H71 treatment, whereas the phosphorylation levels of MAPK p38 remained unaffected. Immunoblotting and phosphoflow assays corroborated the inhibition of AKT and Stat5 signaling by PU-H71 in stem and progenitor populations.

In summary, FLT3-ITD+ AML cells display a stronger response to PU-H71, suggesting that the FLT3-ITD mutation results in a higher dependency on Hsp90 to stabilize the aberrant signaling elicited by constitutive activation of FLT3. Our data suggests that PU-H71 represents a novel therapy for FLT3-ITD+ AML patients with the potential to ablate AML-SCs.