Optimizing Effects of mTOR Inhibition in Acute Myelogenous Leukemia

Introduction: Mammalian target of rapamycin (mTOR) signaling has previously been identified as a possible therapeutic target in acute myelogenous leukemia (AML), as the PI3K/AKT/mTOR pathway has been shown to be upregulated in blasts of up to 90% of AML cases through the constitutive loss of the tumor suppressor PTEN. The activation of this pathway is implicated in synthesis of prosurvival transcription and translational factors responsible for cellular resistance to apoptosis, resistance to chemotherapy, and enhanced survival in the marrow microenvironment. Unfortunately, exploitation of this pathway has been largely ineffective in clinical studies. A second mTOR complex (mTORC2) will reestablish the activation of the first when the first mTOR complex (mTORC1) is shut down under treatment with traditional rapalogs, such as rapamycin and everolimus. Additionally, the cellular process of autophagy has been postulated to be a protective mechanism for leukemic blasts during treatment with mTOR inhibitors, therefore decreasing their efficacy. For these reasons, this work explores means to optimize mTOR pathway inhibition by examining effects of dual mTORC inhibition (OSI027), dual PI3K/mTOR inhibition (BEZ235/Dactilosib), combination of mTOR inhibitors with a histone deacetylase inhibitor with ability to inhibit AKT (LBH589/Panobinostat), and the combination of mTOR inhibitors with hypomethylating agents able to overcome the hypermethylation of mTOR pathway components such as TSC1, TSC2 and PTEN (decitabine/5-azacytidine).

Materials and Methods: Primary AML leukapheresis samples obtained with informed consent were cultured for 48 hours in the presence of OSI027, LBH589 (Novartis), BEZ235 (Novartis), decitabine, or OSI027 in combination with each of these individual inhibitors and compared with control culture conditions. Cells were stained with annexinV and DAPI as previously described, and the percent of living, pre-apoptotic, necrotic and dead cells were determined via flow cytometry. Synergy calculations were completed using Calcusyn™ software as described by Chao and Talalay (1977). Autophagic flux was determined in the OSI027 and LBH589 combination exposure via flow cytometry using anti-LC3. Lastly, protein expression under treatment with various combinations was determined via standard western blotting techniques.

Results: The MV411 AML cell line was utilized to establish efficacy of combination exposures. These cells expressed phosphorylated (p)Akt (ser 473), p4EBP1, and pmTOR. OSI027 10µM, LBH589 1µM, and BEZ235 10nM were able to inhibit pmTOR, pAKT, and p4EBP1 expression as determined by western blotting. In the MV411 cell line, IC50s were determined, and combination indices were determined for OSI027+BEZ235, panobinostat+BEZ235, and panobinostat+OSI027; all of which demonstrated synergy. In primary AML samples, the combination of OSI027 10uM and LBH589 10nM significantly increased the percentage of dead cells in comparison to OSI027 treatment alone. Additionally in primary AML blasts, pmTOR and pAKT expression were decreased in the combination of OSI027 and LBH589 when compared to single agent treatment via western blot. The combination of OSI027 and LBH589 demonstrated a significant reduction in the autophagic flux in comparison to OSI027 treatment alone, suggesting an anti-autophagic effect, which correlates with the increased rates of cellular death in the combination. It was also found in AML cell lines that the combination of 10 µM azacytidine and 100 nM rapamycin resulted in synergistic suppression of U937, MV411, and KG1a survival as measured by MTT. No synergy was noted in the HL60 cell line. This combination effectively suppressed CFU-L in primary AML cases and suppressed p70S6K and p4EBP1 expression more than either agent alone by western blotting. The combination of decitabine 500nM with OSI027 10uM did not demonstrate an increase in apoptotic cells in primary blasts compared to OSI027 alone.

Conclusions: Based on these findings, we conclude that mTOR inhibition in AML cell lines and primary samples can be enhanced through dual mTORC inhibition, PI3K modulation, and histone deacetylase inhibition. This may occur through suppression of activated pathway mediators or through suppression of autophagic flux. These results suggest that there is merit in exploring these combinations for therapeutic potential in AML.