Effective Targeting of Acute Myeloid Leukemia (AML) Harboring the FLT3 ITD Mutation through the Axl/Gas6 Pathway

Abstract 500

Approximately 30% of AML patients harbor activating mutant forms of Fms-like tyrosine kinase-3 (FLT3) such as internal tandem duplications (ITD). To date, targeting FLT3 ITD⁺ AML patients with FLT3 inhibitors has not improved outcomes, suggesting alternative strategies should be pursued. Axl is a type III receptor tyrosine kinase (RTK) which is, along with its ligand Gas6, over-expressed in AML leukemic blasts and correlated with a poor prognosis. In the present study, we asked if blocking the autocrine Axl/Gas6 pathway could impact aberrant proliferation and survival of FLT3 ITD⁺ AML cells. We used a soluble Axl receptor, termed Axl-Fc, to block autocrine binding of Gas6 to Axl. The FLT3 ITD⁺ AML cell line MV4;11 was treated with Axl-Fc in vitro for 4 days and showed a significant reduction in cell proliferation (4.1 ± 0.73 × 10⁶ cells for isotype control (IC) Fc vs. 2.1 ± 0.56 × 10⁶ cells for Axl-Fc, p < 0.05). This growth inhibition resulted from both G1/S cell cycle arrest (51.5 ± 9.5 % at the S phase for IC Fc vs. 2.75 ± 1.5 % for Axl-Fc, p < 0.01), as well as increased apoptosis over control (63.2 ± 8.5 % Annexin V-positive for Axl-Fc vs. 25.7 ± 4.0 % for IC Fc, p < 0.01). Axl-Fc-treated MV4;11 and primary FLT3 ITD⁺ AML blasts showed elevated expression of granulocyte-specific genes compared to IC Fc-treated cells ([MV4;11: lysozyme (48.6 ± 9.5-fold), myeloperoxidise (6.8 ± 0.26-fold), and elastase 2 (2.7 ± 0.45-fold); p < 0.05] [primary FLT3 ITD⁺ AML blasts: elastase 2 (2.4 ± 0.46-fold) and lysozyme (2.7 ± 0.56-fold); p < 0.05]). Morphological analysis showed Axl-Fc could relieve the block in myeloid differentiation for the MV4;11 AML cell line. Treatment with Axl-Fc increased the expression levels of two myeloid transcription factors in MV4;11 (C/EBPα: 16.5 ± 2.6-fold over that seen with IC Fc, p < 0.01; PU.1: 3.9 ± 0.4-fold over that seen with IC Fc, p < 0.01) and in primary FLT3 ITD⁺ AML blasts (C/EBPα: 2.2 ± 0.31-fold over that seen with IC Fc, p < 0.05; PU.1: 2.1 ± 0.36-fold over that seen with IC Fc, p < 0.01). A co-immunoprecipitation study demonstrated that Axl associated with phosphorylated FLT3 in MV4;11 cells. Treatment with Axl-Fc abrogated both the FLT3 phosphorylation and co-precipitation. These in vitro data suggest that the Axl/Gas6 pathway contributes to AML, at least in part, through positively regulating the phosphorylation of FLT3, and that targeting the Axl/Gas6 pathway may diminish phosphorylation of both of these RTKs in AML. In vivo, using a subcutaneous tumor xenograft model, administration of Axl-Fc into MV4;11 tumor-bearing SCID mice significantly suppressed the growth of tumor mass (0.52 ± 0.1 g for PBS-treated mice (n=7); 0.26 ± 0.04 g for IC Fc (n=7) vs. 0.08 ± 0.07 g for Axl-Fc (n=10), p < 0.01). Axl-Fc also diminished the engraftment of primary FLT3 ITD⁺ AML blasts into SCID mice at 8 weeks (WBC: 4.75 ± 0.59 × 10⁶ cells per ml for IC Fc (n=4) vs. 1.6 ± 0.16 × 10⁶ for Axl-Fc (n=5), p < 0.001; 6.9 ± 0.67 % human CD45⁺ cells for IC Fc vs. 2.5 ± 0.59 % for Axl-Fc, p < 0.001). Finally, we assessed Axl-Fc against the FLT3 inhibitor PKC412 (LC Laboratories, Woburn, MA). In vitro treatment of both MV4;11 and primary FLT3 ITD⁺ AML blasts with PKC412 induced phosphorylation of Axl and its downstream signaling molecules, such as ERK and AKT, in both. PKC412-treated MV4;11 cells harboring low level of cleaved caspase 3 (i.e., apoptotic resistant) showed a higher level of phospho-Axl, while MV4;11 cells with high level of cleaved caspase 3 possessed less phospho-Axl (39.6 ± 1.4 % phospho-Axl⁺ for cleaved caspase 3^low population vs. 9.7 ± 3.8 % phospho-Axl⁺ for cleaved caspase 3^high population). These results suggest that activation of the Axl/Gas6 pathway could be, at least in part, responsible for drug resistance against FLT3 inhibition. In in vivo experiment, Axl-Fc was more effective than PKC412 (75 mg/kg) at inhibiting the engraftment of primary FLT3 ITD⁺ AML blasts into SCID mice at 8 weeks (WBC: 2.66 ± 0.36 × 10⁶/ml for PBS (n=5); 1.74 ± 0.19 × 10⁶/ml for PKC412 (n=5) vs. 0.9 ± 0.11 × 10⁶/ml for Axl-Fc (n=5), p = 0.003), while SCID mice treated with Axl-Fc plus PKC412 showed no further reduction of engraftment compared to mice treated with Axl-Fc alone. Survival studies are ongoing. In summary, our study identifies the Axl/Gas6 pathway as an important player for human FLT3 ITD⁺ AML cell growth and survival. Targeting the Axl/Gas6 pathway may prove as effective or superior to small molecule FLT3 inhibitors for the treatment of this subgroup of AML patients.