Interleukin-6 Enhances the Survival of Myeloma Cells by Regulating Bim Binding to Anti-Apoptotic Bcl-2 Proteins

Abstract 4024

The survival of multiple myeloma cells in the bone marrow is largely dependent on its interactions with cellular components of the microenvironment. These interactions, in the form of cell-cell interactions and soluble factors support myeloma cell survival and allow for the mutations associated with transformation to drive plasma cell growth and proliferation. Moreover survival signals from the bone marrow microenvironment influence the response to therapeutic agents, therefore understanding the molecular basis of these interactions is necessary to improve the depth of responses to therapy in myeloma patients.

Cell survival is regulated by the Bcl-2 family of proteins and we previously reported that in myeloma cells one can map which Bcl-2 protein is necessary for cell survival by determining which protein is sequestering the pro-apoptotic molecule, Bim. This could also be used to predict myeloma cell sensitivity to th Bcl-2/Bcl-x_L antagonist ABT-737. While studying the response of freshly isolated patient samples to ABT-737, we found that in 40% of the samples tested (N=10) as well as in 5 myeloma cell lines, inclusion of stromal components significantly inhibited ABT-737-induced apoptosis. Thus the microenvironment can influence survival by altering the pattern of Bcl-2 dependence in cells. We have determined that a soluble factor secreted by stromal cells is necessary for this activity and now report that this factor is interleukin-6.

We found that addition of conditioned medium from the HS-5 stromal line or freshly isolated bone marrow stromal cells (N=4) from myeloma patients or normal donors (N=1) to human myeloma cell lines all resulted significant resistance to ABT-737. Since all of these cells are a rich source of IL-6, we tested if the addition of recombinant IL-6 mimicked the addition of stromal cell supernatant. Indeed rIL-6 (10 ng/ml) was as potent as co-culture with HS-5 cells or addition of HS-5 or stromal cell conditioned medium at blocking ABT-737-induced apoptosis in MM.1s cells. Thus IL-6 is sufficient to block ABT-737-induced apoptosis. To demonstrate that IL-6 was necessary we determined the effect of the addition of an IL-6 neutralizing antibody on stromal cell-induced resistance to ABT-737. We found that addition of the antibody to HS-5 conditioned medium blocked resistance to ABT-737 in a dose-dependent fashion. Moreover we found that addition of anti-IL6 to a patient sample had a similar effect as removing the stromal cells on the response to ABT-737-induced death. Taken together these data suggest that IL-6 is both necessary and sufficient for regulating Bcl-2 dependence in multiple myeloma.

To define the molecular basis for these findings we initially investigated the expression of the key Bcl-2 proteins involved in Bcl-2 dependence and myeloma cell survival. Despite their profound effects on ABT-737-induced apoptosis we found that neither co-culture of MM.1s cells with HS-5 cells, HS-5 conditioned medium or rIL-6 had any effect on the expression of Mcl-1, Bcl-x_L or Bim. We then investigated the patterns of interaction between these proteins and found that the addition of stromal components or IL-6 resulted in a shift of Bim binding from Bcl-x_L to Mcl-1. This is consistent with our previous findings that when Bim is bound to Mcl-1, myeloma cells are less sensitive to ABT-737. Finally we set out to determine how IL-6 signaling results in a change in ABT-737 sensitivity. We treated MM.1s cells with ABT-737 and IL-6 in the presence of inhibitors of the MEK/ERK, p38, PI3K/AKT, and JAK/STAT pathways. We found that only the MEK inhibitor, U0126 could re-sensitize IL-6-treated cells to ABT-737-induced death at concentrations that did not kill cells or have additive effects with ABT-737 in the absence of IL-6 (P<0.05 2-way ANOVA). Taken together, these data indicate that stromal cells regulate myeloma cell survival through IL-6 by altering Bim binding in a MEK-dependent fashion. This results in a cell that is highly dependent on Mcl-1. These data offer an explanation for why proteasome inhibitors are so effective in this disease. While the bone marrow microenvironment is enforcing Mcl-1 dependence, proteasome inhibitors induce the Mcl-1 inhibitor Noxa, which overcomes this activity of IL-6. Moreover these data point to potential ways of combining the targeting of Bcl-2 proteins and IL-6 signaling as a means to effectively kill myeloma cells in the bone marrow microenvironment.