Azacytidine and Gemtuzumab Ozogamicin Co-Operate to Inhibit the Wnt Pathway and Increase Cytotoxicity in AML

Thirty percent of acute myeloid leukaemia (AML) patients contain somatic mutations (internal tandem duplications-ITD) of FLT3 gene (FLT3/ITD) that are associated with poor prognosis. We have previously shown that Gemtuzumab Ozogamicin (Mylotarg), the CD33 calichaemicin-conjugate, targets the leukaemic stem and progenitor cells (LSPC) of FLT3/ITD samples better than WT samples (ASH 2007, Abstract #650). The Wnt signalling pathway has a role in the pathogenesis of AML, particularly in cells capable of self renewal, i.e. LSPC. Beta-catenin is one downstream effector of Wnt signalling, which upon stimulation will activate transcription of target genes, particularly oncogenes such as c-myc and cyclin D1, that influence cell proliferation, survival and cell fate. More recently, a synergistic effect on activation of Wnt signaling pathway in leukaemic transformation by FLT3/ITD has been shown, suggesting that beta-catenin induction may be one transforming event in AML achieved by FLT3/ITD activation. Therefore, identifying therapeutic agents that specifically target the Wnt pathway at various positions may be necessary for complete inhibition of this pathway in AML. In haemopoietic cells, treatment with the hypomethylating agent 5′-Azacytidine (Aza) results in progressive demethylation of previously methylated Wnt inhibitors and re-expression of transcripts and downregulation of Wnt signalling. The aim of this study is to investigate whether Aza targets the Wnt pathway in AML, whether the combination with Mylotarg further frustrates this pathway leading to cell death and whether this is more pronounced in FLT3/ITD samples.

We examined cell survival in the presence of Aza, Mylotarg or the combination in FLT3/WT (HL60 and U937) and FLT3/ITD (MOLM13 and MV4-11) cell lines. Both single Aza and Mylotarg 48hr treatment resulted in significant cell kill (30–42% and 51–74%, respectively), whereas this effect was enhanced (additive) when the Aza/Mylotarg combination was used (69–85% cell kill). Although slightly more sensitive in FLT3/ITD cells (76–85% cell kill), our data indicate that the Aza/Mylotarg combination is effective even in FLT3/WT cells (69–71% cell kill). In order to assess possible mechanism of action, beta-catenin levels were measured in these treated cells. Reduction in beta-catenin levels was similar in Aza-treated FLT3/ITD and WT cells (7–23% and 9–13%, respectively). In contrast, Mylotarg and Aza/Mylotarg combination resulted in a significant beta-catenin reduction in FLT3/ITD cells compared to WT cells: 41–47% in FLT3/ITD and 0–16% in FLT3/WT for Mylotarg and 52–59% in FLT3/ITD and 0–17% in FLT3/WT for Aza/Mylotarg combination. The FLT3 inhibitor AG1296 also downregulated beta-catenin levels in MV4-11 cells, suggesting that Mylotarg may in fact be targeting the Wnt pathway by inhibiting FLT3 signalling.

We have also compared cell survival in FLT3/ITD and compared them to FLT3/WT in LSPC from primary AMLs. CD34⁺CD38⁻CD123⁺ LSPC from FLT3/ITD samples were more sensitive to Aza compared to LSPC of FLT3/WT samples (15% vs 42% cell kill, respectively; n=3, 72hrs). The difference is enhanced when the Mylotarg/Aza combination is used (23% vs 61% cell kill in FLT3/WT and FLT3/ITD, respectively). We conclude that the Mylotarg/Aza combination effectively inhibits the Wnt pathway and targets LSPCs in AML.