The Multitargeted Receptor Tyrosine Kinase Inhibitor Linifanib (ABT-869) Induces Apoptosis through an Akt and Glycogen Synthase Kinase 3β–Dependent Pathway Free

The FMS-like receptor tyrosine kinase 3 (FLT3) plays an important role in controlling the differentiation and proliferation of hematopoietic cells. Somatic mutations in the FLT3 have been frequently identified in acute myeloid leukemia (AML; refs. 1–3). Mutations in FLT3 primarily consist of internal tandem duplications (ITD) in the juxtamembrane domain, affecting 15% to 34% AML patients, or point mutations in the tyrosine kinase domain in 8% to 12% of patients (1, 2). These mutations are associated with a poor prognosis in both adult and pediatric AML patients (1, 2). Mutations result in autophosphorylation of the FLT3 kinase domain and, as a consequence, there is upregulation and activation of downstream signaling pathways such as the Ras/Raf/MEK/ERK pathway, the phosphoinositide-3 (PI3K) kinase pathway (PI3K/PTEN/Akt/mTOR), and the Janus activated kinase (JAK)/STAT pathways (3, 4). Consequently, there is uncontrolled proliferation, arrest of myeloid cell differentiation, and increased resistance to apoptosis.

AML patients receiving conventional chemotherapy regimens experience significant toxicity and relapse due to drug resistance (5). As a result, inhibitors targeting FLT3, with lower toxicity and higher potency than conventional chemotherapy, have emerged and are currently being investigated (5). Preclinical studies using these inhibitors have shown an effect at inhibiting proliferation and inducing apoptosis in human FLT3 mutant cell lines (5). In addition, in vitro studies on the effects of FLT3 inhibitors on human leukemia cell lines with FLT3 mutations have shown inhibition of downstream members of the PI3K pathway such as Akt, members of the Ras/Raf/MEK/ERK pathway such as ERK1/2 and MEK1/2, members of the JAK/STAT pathway such as STAT5, and cell-cycle regulators such as cyclin D, cyclin E, p21^waf1/cip, and p27k^ip1 (6–10). Furthermore, FLT3 inhibitors have been shown to affect members of the Bcl-2 family of apoptotic proteins, such as the proapoptotic proteins BAD and Bim, and antiapoptotic proteins Bcl-xl and Mcl-1 (6, 10–13).

Linifanib (ABT-869) is an ATP-competitive tyrosine kinase inhibitor effective against constitutively active FLT3 and other members of the platelet-derived growth factor receptor (PDGF) and VEGF receptor VEGFR) families (14). Linifanib has been shown, in vivo, to be effective against AML cells harboring FLT3 mutations (MV-411), highly angiogenic fibrosarcoma, small–cell lung carcinoma, epidermoid carcinoma, breast carcinoma, and colon adenocarcinoma (14, 15). Treatment of AML cells with linifanib in combination with other FLT3 inhibitors, such as CEP-701 (lestaurtinib), or chemotherapy, such as cytosine arabinoside (Ara-C) and doxorubicin, have shown synergistic effects (15). Preclinical studies have shown that linifanib inhibits proliferation in FLT3 ITD–positive human leukemia cell lines MV-411 and Molm-14 at a half-maximal inhibitory concentration (IC₅₀) of less than 10 nmol/L (7, 10). Furthermore, linifanib causes cell-cycle arrest and apoptosis through decreased expression of cyclins D and E and increased expression of cyclin-dependent inhibitors p21^waf1/cip and p27k^ip1 (10). In addition, increased expression of proapoptotic BAD, BAK, and BID and decreased expression of antiapoptotic protein Bcl-xl are observed (7, 10, 15). In addition to inhibiting phosphorylation of the FLT3 receptor, linifanib has been shown to have an inhibitory effect on downstream kinases including Akt, ERK, STAT5, and Pim-1 (7, 10).

Many of the previous studies were conducted with human FLT3 ITD leukemia cell lines that may contain other mutations or aberrant signaling pathways. Molecular pathways inhibited by linifanib downstream of FLT3 ITD alone, in the absence of other potential molecular abnormalities, have not yet been studied. To characterize the effects of linifanib specifically on FLT-3 dependent pathways, we used the Ba/F3 pro-B cell line as the model system (16, 17). Modifications to Ba/F3 cells rendering FLT3 receptor constitutively active have shown induction of leukemia-like syndrome in syngeneic mice (18, 19). Ba/F3 pro-B cells require the presence of interleukin-3 (IL-3) to grow and, without it, undergo rapid apoptosis (20, 21). Ba/F3 cells containing the human FLT3 ITD mutation, however, are able to survive independently of IL-3 (18). Phosphatidylinositol 3-kinase (PI3K) and its downstream target, the protein kinase Akt, have an important role in suppressing apoptosis and regulating this growth factor-dependent survival (22–24). In addition, glycogen synthase kinase 3 (GSK3), a serine (Ser)-threonine (Thr) kinase, has been shown to play a role in growth factor withdrawal–induced apoptosis (24, 25). It has been reported that the IL-3 withdrawal–based mechanism of apoptosis was dependent on GSK3-driven mitochondrial outer membrane permeabilization (24). Furthermore, GSK3 has recently been implicated in sustaining proliferation of acute leukemia caused by mixed-lineage leukemia (MLL; ref. 26).

In this study, we show that Ba/F3 pro-B cells with the human FLT3 ITD mutation and treated with linifanib undergo apoptosis and inhibition of proliferation. We show that treatment with linifanib causes reduced phosphorylation of GSK3β at Ser9. This finding is significant, as GSK3β has not been previously characterized to play a role in FLT3 ITD mutant signaling in AML cells and, thus, may play an important role in combining targeted therapies.

Ba/F3 human FLT3 wild type (WT) and FLT3 ITD mutant cell lines were generated by site-directed mutagenesis in the laboratory of Dr. Michael Heinrich (16). Cells were tested and authenticated by Sanger Sequencing of genomic DNA using pLXSN sequencing primers 5′-CCCTTGAACCTCCTCGTTCGACC-3′ and 5′-GAGCCTGGGGACTTTCCACACCC-3′ in 2007. Ba/F3 WT cells were purchased from the American Type Cell Culture (ATCC). Ba/F3 WT and Ba/F3-hFLT3 WT cells were cultured in RPMI-1640 (Gibco/Invitrogen) medium with 10% FBS (Omega Scientific Inc.), penicillin (10,000 units/mL), streptomycin (10,000 μg/mL), l-glutamine (29.2 mg/mL; Gibco/Invitrogen), and 10% conditioned Wehi-3 media, as a source of IL-3. Ba/f3 FLT3 ITD were maintained in RPMI-1640 (Gibco/Invitrogen) with 10% FBS (Omega Scientific, Inc.), penicillin (10,000 units/mL), streptomycin (10,000 μg/mL), and l-glutamine (29.2 mg/mL; Gibco/Invitrogen). Cells were cultured at 37°C and 5% CO₂ in a humidified atmosphere. For experiments involving WT cells grown in human FLT3 ligand, a concentration of 100 ng/mL was used (Sigma) instead of the Wehi-3 supernatant. MV-411 cells were obtained from ATCC and maintained in Iscove's Complete Medium (Gibco/Invitrogen), 10% FBS (Omega Scientific Inc.), penicillin (10,000 units/mL), streptomycin (10,000 μg/mL), and l-glutamine (29.2 mg/mL; Gibco/Invitrogen).

Lyophilized powder of linifanib was provided by Abbott Pharmaceutical, Inc. The compound was dissolved in dimethyl sulfoxide (DMSO; Sigma) to a concentration of 10 mmol/L and stored in aliquots at −80°C for use in in vitro experiments. Linifanib was dissolved in corn oil for in vivo experiments at a concentration of 4 mg/mL and stored in aliquots at −80°C. The structure of linifanib is shown in Supplementary Fig. S1.

To assess cell proliferation, we measured the innate metabolic activity of Ba/F3 FLT3 ITD and Ba/F3 FLT3 WT cells with alamarBlue aqueous dye (Biosource Inc.). We plated 1 × 10⁴ (in 100-μL media) cells at 97% confluency overnight in a 96-well plate (Corning Incorporated); these were treated, the next day, with 100 pmol/L, 1 nmol/L, 10 nmol/L, 100 nmol/L, 1 μmol/L, or 10 μmol/L of linifanib, vehicle control (DMSO), or left untreated. Cell–drug mixtures were allowed to incubate for 24 hours. At 24 hours, 10% alamarBlue dye was added, and plates were incubated for 4 hours. The alamarBlue reduction was measured spectrophotometrically at wavelengths of 540 and 620 nm to calculate the percent reduction of alamarBlue dye as a measure of reduced cell metabolic activity. To assess total cell number, we plated 5 × 10⁴ (in 0.5 mL of media) cells at 97% confluency overnight in a 24-well plate (Corning Incorporated); these were treated, the next day, with 100 pmol/L, 1 nmol/L, 10 nmol/L, 100 nmol/L, 1 μmol/L, or 10 μmol/L of linifanib, vehicle control (DMSO), or left untreated. Cell–drug mixtures were allowed to incubate for 24 hours. Cells were counted using Trypan Blue exclusion and Vi-CELL Cell counter (Beckman Coulter).

Ba/F3 FLT3 ITD mutant cell lines and Ba/F3 FLT3 WT control cell lines (1 × 10⁶ in 10 mL media) were plated overnight in 25-cm² cell culture flasks (Corning Inc.). Cells were either left untreated, treated with vehicle control (DMSO), or 1 nmol/L, 10 nmol/L, 100 nmol/L, 1 μmol/L, or 10 μmol/L concentration of linifanib for 24 hours. After 24 hours, Ba/F3 FLT3 ITD and WT cells (1 × 10⁶) were first washed in cold 1× PBS and then incubated with a 1/400 dilution of Annexin V (Roche) and propidium iodide (PI; Roche) in incubation HEPES buffer (Roche). Data were acquired at the University of California-Los Angeles (UCLA) Johnsson Comprehensive Cancer Center Flow Cytometry Core Facility on the LSR II Analyzer (Becton Dickinson) and analyzed using FlowJo software (www.flojo.com).

For statistical analysis of apoptosis assays, we used AnalystSoft Inc., StatPlus:mac (statistical analysis program for Mac OS; Version 2009; See www.analystsoft.com/en/). Data are presented as the mean ± SD of percentage of Annexin V/PI (early and late apoptosis)-positive cells. Groups were compared using one-way ANOVA, followed by Tukey's HSD post hoc test for differences between means. A value of P < 0.05 was considered to be statistically significant.

Ba/F3 FLT3 ITD mutant cell lines (1 × 10⁶ in 10 mL of media) were plated overnight. Cells were then treated with linifanib (10 nmol/L) alone, and with or without recombinant mouse IL-3 (1 μg/mL; R&D systems) for 24 hours. Cells (1 × 10⁶) were then stained with Annexin-V and PI and analyzed by flow cytometry for apoptosis as described above. A total of 1 × 10⁶ cells were then lysed with radioimmunoprecipitation assay (RIPA) buffer. Whole-cell lysates were run on SDS-PAGE, and Western blot analyses were probed with anti-phospho-GSK3β (Cell Signaling Technology).

Ba/F3 FLT3 ITD cells (3 × 10⁶ in 30 mL of media) were plated on 75-cm² cell culture flasks at a concentration of 1 × 10⁵ cells/mL and incubated overnight. For PARP blots, after overnight incubation, cells were treated for 0, 2, 4, and 6 hours with 1, 10, or 100 nmol/L of linifanib. Cell lysates were run on SDS-PAGE, and Western blot analyses were probed with anti-uncleaved and cleaved PARP antibodies (Cell Signaling Technology).

For FLT3 and Akt blots, cells were treated after overnight incubation for 15, 30, 60, and 120 minutes with linifanib (10 nmol/L) or vehicle control (DMSO). Cell lysates were immunoprecipitated with 1 μg/mL of anti-FLT3 antibodies (Santa Cruz Biotechnology) or anti-Akt (Cell Signaling Technology) antibodies and Protein A/G Sepharose beads (Santa Cruz Biotechnology). Western blot analyses were probed with (i) anti-phospho-FLT3 antibody Tyr591 (Cell Signaling Technology) or (ii) anti-phospho-Akt (Ser473) antibodies (Cell Signaling Technology).

For GSK3 immunoblots, cells were treated for 15, 30, 60, and 120 minutes with 10 nmol/L linifanib or with vehicle control (DMSO). Cells were then lysed with RIPA buffer. Whole-cell lysates were run on SDS-PAGE, and Western blot analyses were probed with anti-phospho-GSK3β (Cell Signaling Technology) or anti-GSK3 (α/β; Cell Signaling Technology). MV-411 cells were treated with 10 nmol/L of linifanib for 1 hour. Cells were then lysed with RIPA buffer. Whole-cell lysates were run on SDS-PAGE, and Western blot analyses were probed with anti-phospho-GSK3β antibodies (Cell Signaling Technology) or anti-GSK3 (α/β; Cell Signaling Technology).

The green fluorescent protein (GFP)-luciferase retrovirus (#74 RRL.sin.cPPT.fLuc.IRES.emdGFP) was obtained from the virus vector core laboratory at UCLA. One day before infection, Ba/F3 WT and Ba/F3 FLT3 ITD mutant cells were diluted to a concentration of 0.5 × 10⁶ cells/mL. Concentrated virus was diluted 1:4 to obtain a concentration of 20 μg/mL. Cells (0.5 × 10⁶) were centrifuged, and 1 mL of diluted virus was added together with 1 mg/mL protamine sulfate (UCLA). The cell–virus suspension was transferred to a 6-well plate and incubated at 37°C in 5% CO₂ overnight. A day after transduction, cells were washed twice with 2 mL of RPMI media, and 2 mL of Fresh media was added to cell wells. Three days after transduction, cells were sorted for GFP-positive cell population at the UCLA flow cytometry core laboratory.

Non-obese diabetic (NOD)/severe combined immunodeficient (SCID) mice (Jackson Laboratories) received Ba/F3 FLT3 WT, FLT3 ITD GFP-luciferase mutant cell lines (1 × 10⁶; 5 mice per group) by tail vein injection. Mice were then treated by daily oral gavage with 0.2 mL/20 gm mouse weight with linifanib or with vehicle control (corn oil). Mice were monitored for disease progression by measurement of weight loss and bioluminescence imaging. Bioluminescent images were acquired using Xenogen IVIS hardware (Xenogen) and Living Image software (Caliper Life Sciences). Before acquisition, mice were anesthetized with isoflourane and subsequently given 126 mg/g mouse weight of d-luciferin by intraperitoneal (i.p.) injection for detection of luciferase. Animals were sacrificed after they showed symptoms of illness such as ruffled fur, labored breathing, and hunched back.

Survival data were analyzed using the SAS program (SAS Institute; www.sas.com) and a Kaplan–Meier survival model. The log-rank test was used for comparing survival curves.

To determine whether linifanib had antiproliferative and apoptotic effects in vitro on ITD mutant cell lines, we conducted dose–response alamarBlue assays and apoptotic assays on both Ba/F3 FLT3 ITD mutant and WT cells. The assays show that, after 24 hours, linifanib is more effective at inhibiting cell growth in ITD mutant cells than in WT cells (Fig. 1A). The IC₅₀ of linifanib on ITD cells was 0.55 nmol/L, whereas the IC₅₀ for WT cells was 6 μmol/L (Fig. 1A). Culturing WT cells with FLT3 ligand, however, showed similar inhibition of cell growth as in ITD mutant cells; minor differences can be accounted for by differences in rate of cell growth (Fig. 1A and Supplementary Fig. S2). This showed that the effects of FLT3 inhibitor were specific to FLT3. Furthermore, viable cell counts were measured (Supplementary Fig. S2). In addition, treatment with 10 nmol/L of linifanib induced apoptosis in ITD mutant cells, whereas no effect was observed on WT cells (Fig. 1B). Linifanib treatment did not show any differences at reducing cell viability or inhibiting proliferation between WT and FLT3 mutant cells containing the D835V point mutation (data not shown). To ascertain the time frame for induction of apoptosis, we treated ITD mutant cells with linifanib in a time course from 0 to 24 hours. PARP cleavage was detected as early as after 6 hours of treatment (Fig. 1C).

In vivo, xenograft experiments with NOD/SCID mice showed that mice injected with ITD mutant cells and treated daily orally by gavage with linifanib had a decreased rate of leukemia progression compared with untreated mice (Fig. 2). On day 7, untreated mice showed rapid progression of ITD mutant cells, whereas mice treated with linifanib had no detectable disease on testing by bioluminescence (Fig. 2A). In addition, survival duration of untreated mice receiving ITD mutant cells was significantly shorter (P < 0.01) than for those receiving daily treatment with linifanib or injected with WT cells (Fig. 2B).

As linifanib showed antiproliferative and apoptotic effects on ITD mutant cells both in vitro and in vivo, we next sought to examine the mechanism by which these occurred.

Because treatment with linifanib has been shown to induce apoptosis rapidly (within 6 hours), we hypothesized that apoptosis induced by linifanib results from Ba/F3 FLT3 ITD mutant cells defaulting to an IL-3–deficient state and, thereby, undergoing apoptosis. Therefore, we hypothesized that adding IL-3 would reverse linifanib-induced apoptotic effects.

To test this hypothesis, recombinant IL-3, in combination with 10 nmol/L linifanib, was added to cells (Fig. 3). Our data revealed that adding recombinant IL-3 reversed the apoptotic effects of linifanib alone with a reduction from 40.2% overall apoptosis with linifanib treatment alone down to control levels (P < 0.001; Fig. 3).

IL-3 withdrawal–induced apoptosis has been shown to occur through the PI3K/Akt/GSK3β pathway (24). Because ITD mutant cells were rescued with IL-3, we hypothesized that linifanib works through the same pathway. To test this possibility, we next sought to determine whether PI3K, Akt, and GSK3 are downstream kinase targets affected by treatment with linifanib.

It has been established that, in the IL-3–dependent cells, removal of IL-3 induces apoptosis by inhibiting Akt and GSK3 phosphorylation (22–24). Because IL-3 rescues linifanib-induced apoptosis, we hypothesized that treatment with linifanib reduces phosphorylation of Akt and GSK3 in the Ba/F3 FLT3 ITD mutant cell line.

To test this possibility, ITD mutant cell lines were examined for phosphorylation of Akt and GSK3β by immunoprecipitation, SDS-PAGE, and Western blot analysis (Fig. 4). We show that linifanib is effective at inhibiting phosphorylation of FLT3 in Ba/F3 FLT3 ITD cell lines at a concentration of 10 nmol/L (Fig. 4A). In addition, linifanib reduced phosphorylation of Akt at Ser473 after treatment with 10 nmol/L linifanib (Fig. 4B). To test whether GSK3β phosphorylation was affected after treatment with linifanib, we treated the ITD mutant cells with 10 nmol/L linifanib and examined phosphorylation of GSK3β at Ser9 (Fig. 4C) or GSK3α at Ser21 (data not shown). Treatment with 10 nmol/L linifanib resulted in decreased phosphorylation of GSK3β Ser9 as early as 60 minutes (Fig. 4C). GSK3α at Ser21 only showed reduced phosphorylation after 8 hours (data not shown). To test whether GSK3β phosphorylation is rescued with recombinant IL-3, we treated the ITD mutant cells with a combination of 10 nmol/L linifanib and recombinant IL-3 and, then, examined phosphorylation of GSK3β at 24 hours. Treatment with a combination of linifanib and IL-3 resulted in rescue of GSK3β phosphorylation (Fig. 4D). To test whether the same GSK3β phosphorylation is observed in human AML FLT3 ITD mutant cells, the MV-411 cell line was treated with linifanib. It was found that treatment with 10 nmol/L of linifanib reduced GSK3β phosphorylation as well (Supplementary Fig. S3). This emphasizes the importance of GSK3β in not only mouse cells but also human cells.

Our results, therefore, suggest that one of the possible mechanisms by which linifanib induces apoptosis is through modulation of Akt and GSK3β phosphorylation.

To determine whether GSK3 has a major role in inducing apoptosis on treatment with linifanib, we treated ITD mutant cells with a combination of 10 nmol/L linifanib and 10 mmol/L lithium chloride, a known GSK3 inhibitor (Fig. 5). We hypothesized that, because GSK3β phosphorylation is reduced as a result of linifanib treatment, it may have a major role to play in induction of apoptosis in ITD mutant cells. Although not as large as we expected, we have shown that combination treatment with lithium chloride causes a reduction in apoptosis at 24 and 48 hours (P < 0.01;Fig. 5). These results suggest that modulation of GSK3β phosphorylation may be at least a contributing factor for linifanib-induced apoptosis.

In this article, we have characterized a new downstream target of linifanib-induced FLT3 inhibition. We have shown that FLT3 inhibition by linifanib in ITD mutant cells results in reduced GSK3β phosphorylation.

Initially, we showed that linifanib rapidly induces apoptosis in ITD mutant cell lines. Because of this, we hypothesized that linifanib induces apoptosis in ITD mutant cells by mimicking IL-3 withdrawal–induced apoptosis. Therefore, we speculated that IL-3 would rescue any linifanib-induced apoptotic effects. Our data have shown that IL-3 is able to reverse the effects of linifanib-induced apoptosis.

Furthermore, we hypothesized that, because IL-3 rescues the effects of linifanib-induced apoptosis, that apoptosis in ITD mutant cell lines occurs through the same pathway as IL-3 withdrawal–induced apoptosis by inhibiting PI3K activation, reducing Akt phosphorylation, and reducing phosphorylation of GSK3β. Our data have shown that treatment with linifanib reduces Akt phosphorylation and GSK3β phosphorylation.

Other studies with FLT3 inhibitors have shown that inhibiting FLT3 phosphorylation leads to suppression of downstream targets, such as STAT5, members of the PI3K pathway, mitogen-activated protein kinase pathway, and the Bcl-2 family of proteins, and cell-cycle regulators (6–10, 13). As seen in previous studies in ITD mutant cells, we have observed similar downstream targets of linifanib including Akt, ERK1, Bcl-xl, and BAD (data not shown). However, GSK3β as a target of linifanib has not yet been characterized.

GSK3 is a Ser-Thr protein kinase that regulates cell differentiation and apoptosis, the canonical Wnt signaling pathway, as well as glycogen synthesis (27). GSK3 has been shown to phosphorylate substrates such as cytoskeletal proteins, affect cell-cycle regulation by targeting β-catenin, MYC, cyclin D1, cyclin E, and Bcl-3, transcription factors such as c-Jun, c-Myc, c-Myb, and cAMP-responsive element-binding protein, and other metabolic regulators (24, 28, 29). Although increased activity of GSK3 has been observed in chronic metabolic disorders such as type II diabetes, mood disorders, Alzheimer's disease, and in acute leukemia caused by MLL, its role has not yet been characterized in AML with FLT3 ITD mutations (25, 26, 28).

In growth factor–dependent hematopoietic cells, it has been shown that one of the pathways responsible for survival is the PI3K/Akt pathway (30). In addition, dominant-negative forms of Akt were able to accelerate IL-3–induced apoptosis (30). Furthermore, recent studies have shown that growth factor–induced apoptosis occurs by reducing phosphorylation of GSK3β (reducing its kinase activity; ref. 24). In addition, it has been shown that inhibiting GSK3β activity through a variety of small-molecule inhibitors prevented apoptosis from occurring (24). We propose that Ba/F3 FLT3 ITD mutant cell lines are able to survive in an IL-3–independent manner because the FLT3 ITD constitutive mutation renders these cells alive through PI3K/Akt signaling, which is the same pathway as that of IL-3 survival (Fig. 6A; ref. 24). However, we propose that inhibiting FLT3 with linifanib prevents PI3K activation, reduces Akt and GSK3β phosphorylation, and, therefore, ITD mutant cell lines default to a mechanism mimicking IL-3 withdrawal–induced apoptosis (Fig. 6B and C). Studies with one other FLT3 inhibitor, AG1296, also saw similar rescue of apoptosis by IL-3, but the role of GSK3β was not characterized in this study (31).

Further studies are required to understand the precise role of GSK signaling in the pathogenesis of AML cells. Commercially available GSK3 inhibitors could be used to characterize these pathways. Our preliminary studies using the lithium chloride inhibitor found a slight reduction in overall apoptosis when combined with linifanib (Fig. 5). There is evidence that GSK3 does have a role in linifanib-induced apoptosis, although it may not be the only factor involved in inducing apoptosis in the ITD cells, as there may be cross-talk between other pathways downstream of FLT3 activation that can also affect apoptosis. Signaling targets such as GSK3β, however, may help to elucidate the mechanism by which linifanib induces apoptosis. Combination studies of FLT3 inhibitors with other inhibitors have been successful at inhibiting the progression of AML by enhancing apoptosis and antiproliferative effects (32, 33). GSK3 inhibitors may be alternative viable candidates in these combination studies.

In conclusion, the development of FLT3 inhibitors for the treatment of AML has been successful to an extent. Previous studies have found that the use of FLT3 inhibitors in conjunction with other inhibitors or with conventional chemotherapy drugs may prove to be more successful in effectively treating AML. The development of drug resistance in human AML cell lines after initial therapy provides an avenue for testing combinations of new inhibitors that target different pathways. The use of FLT3 inhibitors in combination with GSK3 inhibitors or chemotherapy may be a more optimal approach to treat AML.

K.M. Sakamoto received a commercial research grant from Abbott Laboratories, Inc., and received other commercial research support from Genentech, Inc. M.C. Heinrich has equity interest in Molecular M.D.

The authors thank Jenny Chang and Salemiz Sandoval for their advice and assistance with animal work, and Drs. Linda Baum and D.H. Davies for critical review and advice in the preparation of the manuscript.

K.M. Sakamoto is supported by NIH grants HL75826 and HL83077, the William Lawrence and Blanche Hughes Foundation, and the St. Baldrick's Foundation. K.M. Sakamoto is a scholar of the Leukemia and Lymphoma Society. M.C. Heinrich is supported, in part, by a Merit Review grant from the Department of Veterans Affairs and a SCOR grant from the Leukemia and Lymphoma Society.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.