Imipramine Blue Sensitively and Selectively Targets FLT3 and c-Kit Mutant Acute Myeloid Leukemia

Acute myeloid leukemia (AML) is the most common acute leukemia in adults and the 2^nd most common in children. Despite advances in our understanding of AML biology, long term survival remains suboptimal. Current therapy primarily involves high dose cytotoxic chemotherapy and possible allogeneic stem cell transplant, both of which are heavily myelosuppressive and carry some treatment related mortality. The FLT3/ITD mutation is found in 20-30% of adults and 10-20% of children with AML, and despite intensive therapy survival outcomes are dismal. Recent strategies using targeted therapies (e.g. FLT3 tyrosine kinase inhibitors-TKIs) have so far yielded modest responses, and most relapsed/refractory patients will still die of their leukemia. Thus selective agents targeting FLT3 mutated AML but not benign hematopoiesis are needed. A large portion of AML, including those with FLT3/ITD, has elevated levels of reactive oxygen species (ROS), and AML with high ROS is highly proliferative. Interestingly, mislocalization of mutant type III receptor tyrosine kinases (RTKs, e.g. FLT3 and c-KIT) occurs in leukemia. In FLT3/ITD, a portion of underglycosylated RTK accumulates on the endoplasmic reticulum (ER) in a microenvironment with high levels of localized ROS generated from ER-bound NADPH Oxidase 4 (Nox4) and through tight connections with mitochondria. ROS has been demonstrated in recent years to be an integral part of cell signaling, typically through suppression of protein tyrosine phosphatases. Therefore, we hypothesized that elevated ROS plays an important role in leukemic cell signaling in FLT3/ITD+ AML. To this end, we have investigated the effects of the Nox4 inhibitor imipramine blue (IB), a novel potent agent from the class of triphenylmethane dyes (Munson et al, Sci Transl Med, Vol. 4, 2012). We used FLT3/ITD+ AML cell lines (MV4-11, MOLM-14), cell lines without FLT3/ITD (OCI-AML3, K562, HEL, HL-60), and Ba/F3 and 32D cells transduced with FLT3 and c-KIT mutants, respectively. We screened inhibitors of mitochondrial complex I and NOX inhibitors to determine their effects on ROS levels by H2-DCF-DA staining/flow cytometry and on cell viability by trypan blue exclusion assay. While all mitochondrial and NOX4 inhibitors tested effectively reduced total cellular ROS levels, only Nox4 inhibition with IB achieved selective cytotoxicity for FLT3/ITD+ cell lines with high potency (IC₅₀ =125 nM in MV4-11). The Annexin V assay showed that IB causes cell death through an apoptotic mechanism in MV4-11 cells. To determine effects of Nox4 inhibition on leukemic cell signaling, specifically on protein tyrosine phosphatases (PTPs), we studied the signal transducer and activator of transcription 5 (STAT5). We observed phosphorylated STAT5 (pSTAT5) levels by Western blot, and sodium orthovanadate (Na3VO4) was used for PTP inhibition. IB caused dose dependent decreases in pSTAT5 levels after 4 hours of exposure, and this effect was reversed when Na3VO4 was added, implicating reactivated PTPs as the cause of pSTAT5 suppression. To determine the downstream effects of STAT5 inhibition by IB, STAT5 target genes were evaluated by qRT-PCR. Four hour IB treatment caused decreased expression of common STAT5 target genes (Pim1, c-Myc, Cish) indicating effective suppression of STAT5 function.  To optimize cytotoxicity, we then combined IB with the STAT5 inhibitor pimozide. A combination of 75 nM IB/5µM pimozide was highly synergistic in MV4-11 (5 fold higher cytotoxicity over individual drugs, 3 fold higher than their additive effect). IB/pimozide selectively killed FLT3/ITD+ AML cell lines while sparing AML cells without FLT3/ITD and benign CD34+ cord blood cells. IB and IB/pimozide were also equally effective in killing Ba/F3 FLT3/ITD cells with point mutations causing resistance to TKIs and selectively killed 32D cells with c-KIT (D814V) mutation over those with wild type c-KIT. Thus we have shown that the Nox4 inhibitor IB is a potent and selective inducer of apoptosis for FLT3/ITD+ AML in vitro, and the novel combination of IB and pimozide selectively targets FLT3/ITD and c-KIT positive mutants, including TKI-resistant FLT3/ITD mutants. We propose that IB alone or in combination with pimozide represents a novel therapeutic strategy, and testing in primary patient samples and in vivo models are currently underway.