Glycosylation and Surface Localization Are Required for FLT3 Activation but Not for FLT3/ITD.

Abstract 2748

Poster Board II-724

FLT3 is a class III receptor tyrosine kinase that is normally activated on the surface of hematopoietic cells by binding of FLT3 ligand (FL). After FL binding, phosphorylation leads to activation of multiple signaling pathways and to down-regulation of surface expression due to internalization of the receptor. Activating mutations of FLT3 are commonly observed in acute myeloid leukemia (AML) and are associated with a poor prognosis. About 33% of AML cases contain constitutively activated FLT3 either as point mutations in the kinase domain or as internal tandem duplications (ITDs) in or near the juxtamembrane domain. These mutant forms of FLT3 do not require FL and localize both to an intracellular compartment as well as to the plasma membrane. We identified two subclones of TF-1/ITD cells that expressed almost exclusively the intracellular form of FLT3/ITD. These cells were resistant to inhibition by tyrosine kinase inhibitors (TKIs) in the absence of any resistance conferring mutation within FLT3 itself. MTT assays revealed that the IC₅₀ values increased by ∼10-fold for PKC412 in clones TF-1/ITD P8 (P8) and TF-1/ITD P65 (P65) compared to parental TF-1/ITD cells. A high level of pan-resistance was noted for lestaurtinib, sunitinib, sorafenib and AGL2043. In agreement with the MTT data, Western blotting shows that FLT3/ITD in both P8 and P65 was not inhibited by any of the tested TKIs. Blots show that inhibition of FLT3/ITD in TF-1/ITD cells led to formation of more of the fully glycosylated mature receptor. Immunofluorescence microcsopy shows that FLT3/ITD inhibition was accompanied by translocation to the plasma membrane. Unlike the FLT3 TKIs, the src inhibitor PP2 inhibited FLT3 phosphorylation in the P8 and P65 clones. However, FLT3/ITD inhibition by PP2 in P8 and P65 did not promote receptor maturation. Immunofluorescence microscopy shows that wild-type FLT3 resides primarily in the plasma membrane and FLT3/ITD in TF-1/ITD cells that has been inhibited translocates from the trans-Golgi network to the plasma membrane. FLT3/ITD in clones P8 and P65 appear to be restricted to the endoplasmic reticulum, and inhibition by PP2 did not cause translocation to the surface. We wanted to test whether FLT3 with modified glycosylation could mimic the phenotype seen in the P8 and P65 clones. Cells treated with the deglycosylation agents, tunicamycin or swainsonine, produced altered glycosylation patterns of FLT3 and FLT3/ITD. FLT3 could not be stimulated by FL following treatment with tunicamycin, but it could still be stimulated by FL after partial deglycosylation by treatment with swainsonine. These results indicate that FLT3 need not be in its mature form to be fully activated by FL, but earlier steps during glycosylation are required for activation. On the other hand, FLT3/ITD appears to be activated soon after translation and prior to glycosylation as evidenced by its phosphorylation in the absence of even the 14-sugar core oligosaccharide whose transfer was prevented by tunicamycin. It has been reported that localization of receptor tyrosine kinases may affect downstream signaling, but we did not observe noticeable changes in Stat5 or MAP Kinase activation in FLT3/ITD cells treated with tunicamycin or swainsonine. (In one published study, a FLT3/ITD that was altered to anchor to the E.R. showed an increase in Stat3 activation but a decrease in Stat5, Erk and Akt activation.) Thus, it is likely that there are some components of receptor trafficking and/or localization that contribute to transformation phenotypes. Aberrant intracellular localization has been previously documented for the PDGF and Kit receptors in a human glioblastoma and gastrointestinal stromal tumor, respectively. This suggests that altered localization may also contribute to FLT3 transformation phenotype and may affect inhibitor resistance as well.