Identification of a New AHI-1-Mediated-PP2A-Beta-Catenin-BCR-ABL-JAK2 Complex Modulating Resistance of CML Stem/Progenitor Cells to Tyrosine Kinase Inhibitors

Tyrosine kinase inhibitor (TKI) therapies have had a major impact on treatment of early phase CML patients.  However, TKI monotherapies are not curative and initial and acquired resistance remain challenges. Particularly, leukemic stem cells (LSCs) are less responsive to TKIs and are a critical target population for TKI resistance. We previously demonstrated that Abelson helper integration site-1 (AHI-1) is highly upregulated in LSCs and interacts with multiple kinases, including BCR-ABL and JAK2. We further showed that AHI-1-mediated complexes contribute to enhanced transforming activity of BCR-ABL and drug resistance of LSCs, suggesting AHI-1 as a new therapeutic target in CML. By screening the Prestwick Chemical Library, we have recently identified a specific growth inhibitory compound that potentially targets AHI-1: Cantharidin (CAN), an inhibitor of protein phosphatase 2A (PP2A).  PP2A is a family of serine/threonine phosphatases that regulate cell signaling cascades involved in proliferation and cell cycle control of cancer cells. Evidence suggests the dysfunction of specific PP2A protein complexes is primarily responsible for the changes that lead to cell transformation.  It was also reported that PP2A activity is downregulated in CML cells, particularly in blast crisis CML, due to overexpression of SET, and that the use of a PP2A activator inhibits the growth of CML cells. Interestingly, we have now demonstrated that CAN inhibits the growth of AHI-1-transduced BCR-ABL⁺ cells by about 30% compared to control cells, but this effect was significantly enhanced to 93% with the addition of imatinib (IM). As well, AHI-1-suppressed cells were more sensitive to CAN treatment. PP2A is highly expressed in K562 and IM-resistant K562 cells and combination treatment with CAN and TKI prevents growth and induces apoptosis in these cells more effectively than single treatments (2-3 fold, p<0.001). The combination also greatly reduced colony formation (CFC) of CD34⁺ CML cells, but CAN also inhibited CFCs of CD34⁺ normal bone marrow (BM) cells. To overcome toxicity issues, we have recently evaluated new, pre-clinically validated PP2A inhibitors, LB100 and LB102. These specifically suppress PP2A activity up to 75% in CML cells, as determined by immunoprecipitation (IP) phosphatase assays, and inhibit growth of these cells. Importantly, PP2A inhibitors plus TKIs significantly reduced the yield of colonies from CD34⁺ CML stem/progenitor cells compared to any single agent or combination of TKIs and this enhanced effect was most pronounced on cells from IM nonresponders (p<0.05). New PP2A inhibitors are also much less toxic to CD34⁺ normal BM cells compared to CAN (2-3 fold, p<0.001).  Long-term culture-initiating cell assays showed that more primitive cells were eliminated to a greater extent by combination treatments. Moreover, the combination resulted in strong synergy in IM-insensitive cells (CI value <0.5). Notably, cell cycle analysis showed that treatment with LB100/LB102 alone induces a shift from G1 to G2/M phase (3-fold, p<0.05). A similar shift in the cell population was observed after combination treatment with IM, suggesting that the G2/M phase arrest is solely due to PP2A inhibition. Confocal microscopy showed the G2/M arrest, leading to mitotic catastrophe. Mechanistically, we have identified the PP2A subunit B (PR55a) as an AHI-1 interacting protein using IP/mass spectrometry; this was further confirmed in AHI-1-transduced cells by IP-Western analysis. Most interestingly, Western blot analysis showed that the combination treatment not only inhibits tyrosine phosphorylation of BCR-ABL, JAK2, STAT5, AKT and P38, but also significantly inhibits protein expression of these proteins compared to single agents. Inhibition of protein expression of AHI-1 and β-catenin by the combination was also observed and a new protein interaction between AHI-1 and β-catenin was identified.  Moreover, the combination directly affects PP2A-mediated β-catenin dephosphorylation and BCR-ABL-mediated phosphorylation of β-catenin in CML cells, which leads to protein degradation of multiple proteins. In summary, we have uncovered new AHI-1-PP2A (PR55a) and AHI-1- β-catenin interactions and simultaneously targeting both BCR-ABL and PP2A activities in CML LSCs may offer an important new therapeutic possibility, through destabilization of the protein-protein interactions mediated by AHI-1.