A Selective PIM Kinase Inhibitor Is Highly Active In Multiple Myeloma: Mechanism of Action and Signal Transduction Studies

Abstract 4084

Related work from our group has shown the therapeutic utility of PIM inhibition in multiple myeloma cell lines, xenografts, and primary patient samples (Ebens A. et al., ASH 2010 submitted abstr.). In this study we provide detailed mechanistic findings to show that PIM kinase inhibition co-regulates several important elements of the PI3K/AKT/mTOR pathway, resulting in significant synergy for combination drug treatments. The PIM kinases are a family of 3 ser/thr growth factor- & cytokine-induced proteins hypothesized to have redundant survival and growth functions. GNE-652 is a pan-PIM kinase inhibitor with picomolar biochemical potencies and an excellent kinase selectivity profile. Myeloma cell lines exhibit sensitivity to single agent PIM inhibition and a striking synergy in combination with the PI3K inhibitor GDC-0941. Cells respond to this combination with cell cycle arrest and marked apoptosis in vitro. We tested a panel of selective PI3K/AKT/mTOR inhibitors and found PI3K and AKT inhibitors showed the greatest extent of synergy with GNE-652, whereas mTOR inhibitors were synergistic to a lesser extent. These results suggest that PIM signaling converges on both TORC1 and AKT to generate these differential synergies. BAD is a negative regulator of both Bcl-2 and Bcl-XL, and we were able to confirm previous reports that AKT and PIM cooperate to inactivate BAD (Datt et al., 1997; Yan et al., 2003). Pim has been shown to potentially inactivate PRAS40, a negative regulator of TORC1 (Zhang et al., 2009). We demonstrate that PIM or PI3K inhibition caused a loss of phosphorylation on PRAS40 and results in a physical association of PRAS40 and TORC1 and a decrease in phosphorylated p70S6K and S6RP. These reductions were apparent in 7 of 7 cell lines assayed and enhanced by the combination of PI3K and PIM inhibition in these cell lines. Consistent with prior reports (Hammerman et al., 2005), we show that a second node of convergence between PIM and TORC1 is 4E-BP1. Both GDC-0941 and GNE-652 treatments reduced phosphorylation of 4E-BP1 in 7 of 7 myeloma cell lines. Since dephosphorylated 4E-BP1 competes with eIF4G for the mRNA cap binding factor eIF4E, we assayed immunoprecipitates of eIF4E for the presence of eIF4G and 4E-BP1 and observed increased BP1 and decreased 4G. The combination treatment significantly enhanced the loss of 4G relative to either single agent, and importantly, even at 5× the IC50 concentrations for single agents, combination drug treatment achieved greater extent of effect than single agent treatment. Thus PI3K and PIM pathways are redundant at the level of cap-dependent translational initiation mediated by eIF4E. It has been hypothesized a subset of mRNAs are particularly sensitive to inhibition of cap-dependent translation, and that this includes a number of oncogenes such as cyclin D1. We assayed global protein synthesis in MM1.s cells using 35S-methionine and as expected we observed only a modest ≂∼f20% decrease caused by either GNE-652 or GDC-0941 and this decrease was not enhanced by combination treatment. However, we noted across 7 different myeloma cell lines, strong decreases in levels of cyclin D1 that were enhanced by combination treatment. In summary, we have identified several points at which PIM and PI3K/AKT/mTOR converge to provide synergistic apoptosis in multiple myeloma cell lines. These results provide the rationale for further preclinical development of PIM inhibitors and provide the basis for a possible clinical development plan in multiple myeloma.