Phosphorylation Influences The Binding Of Bim To Anti-Apoptotic Proteins In Multiple Myeloma

Multiple myeloma cells are strongly dependent on the anti-apoptotic protein Mcl-1 for survival. However, several studies have shown that some myeloma cells are sensitive to treatment with ABT-737 and related compounds, which target Bcl-2 and Bcl-x_L, but not Mcl-1. These studies suggest that sensitive cells have more of the direct-activating, pro-apoptotic protein Bim bound to Bcl-2 and Bcl-x_L than to Mcl-1. We have previously shown that the expression levels of pro- and anti-apoptotic proteins have little influence on which anti-apoptotic protein binds Bim, leading us to further explore what factors influence the preferential binding of Bim.

We have found that IL-6 stimulation of myeloma cells results in both the phosphorylation of Bim at serine 69, and a shift in Bim from Bcl-2 and Bcl-x_L to Mcl-1. These observations suggest a role for phosphorylation in determining the anti-apoptotic protein that Bim binds to. To determine if constitutive phosphorylation of Bim regulates the steady state binding of Bim to anti-apoptotic Bcl-2 family members, we utilized phospho-affinity gel electrophoresis (Phos-Tag). We determined that Bim is constitutively phosphorylated in unstimulated myeloma cells. However, unlike what is observed following IL-6 stimulation, phosphorylation does not occur at serine 69. Through use of co-immunopreciptation studies, we found that phosphorylation status of Bim affects its binding to anti-apoptotic proteins. The phosphorylated species are preferentially bound to Bcl-2 and Bcl-x_L, and not to Mcl-1. Additionally, we observed that phosphorylated Bim species bound to Bcl-2 and Bcl-x_L were present at a significantly lower level in myeloma cells that were more dependent on Mcl-1, such as OPM2, than in KMS18 cells, which exhibited greater Bcl-2/Bcl-x_L dependency. This pattern also held true for myeloma cells that had been selected for resistance to ABT-737. ABT-737 resistant KMS18 cells had no detectable phosphorylated species of Bim, and all of the Bim was bound to Mcl-1. We have begun to identify the phosphorylated amino acids in Bim through the use of phospho-Bim-specific antibodies. In addition to serine 69, we have determined that the constitutive phosphorylation does not occur at serine 59 or serine 87. However, preliminary data suggests that Bim phosphorylated at threonine 116 preferentially binds to Bcl-x_L in KMS18 cells.

We have also taken initial steps to identify signaling pathways and kinases that are phosphorylating Bim, and altering its affinity. We treated KMS18 myeloma cells for 12 hours with 10 μM of the MEK inhibitor U0126, the JNK inhibitor SP600125, or the p38 inhibitor SB203580, and ran lysates from each of these conditions on PhosTag gels. While U0126 and SB203580 treatment did not result in the loss of any phosphorylated Bim species, treatment with SP600125 resulted in a loss of phosphorylated Bim species comparable to treatment of myeloma cell lysates with lambda protein phosphatase. These results suggest that JNK mediates phosphorylation of Bim in myeloma cells. Interestingly, threonine 116 is contained within a JNK consensus sequence. Therefore, JNK could directly influence the preferential binding of Bim. Together, these results suggest that combining Bim kinase inhibitors with agents that result in decreased Mcl-1 (e.g. CDK9 inhibitors) or inducers of the Mcl-1 inhibitor Noxa (proteasome inhibitors) represents a novel approach in cancer cells that demonstrate Mcl-1 dependence.