Induction of Bim-Dependent and -Independent Apoptosis in Multiple Myeloma

Multiple myeloma is a plasma cell malignancy, estimated to affect 24,050 new Americans in 2014. While new treatments have emerged in recent years greatly increasing survival rates, our understanding of the mechanisms whereby these therapies kill myeloma cells is still lacking. By understanding how drugs induce cell death we can help to predict cell sensitivity, dissect resistance pathways, and design more effective combination treatment regimens.

The Bcl-2 family of proteins regulates cell survival and death. We have previously reported that in myeloma cells, one can map the Bcl-2 protein that is necessary for survival by determining which anti-apoptotic protein is sequestering the pro-apoptotic protein, Bim. Bim is known as a crucial component in the activation of pro-apoptotic proteins Bak and Bax, which is necessary for cells to undergo apoptosis. While other proteins, such as Puma and tBid, are also capable of activating Bak and Bax, Bim is thought to be the primary activator in myeloma cells.

Given the role of Bim in apoptosis, we were interested in determining the contribution of Bim induction in response to therapeutic agents and resistance mechanisms. RNAseq analysis of 69 myeloma cell lines revealed that only two human myeloma cell lines, LP-1 and KMS26 were devoid of Bim expression. This is due to biallelic deletion in both lines as determined by array CGH. Therefore we examined the responses of these two cell lines to ABT-737, which functions by directly inhibiting Bcl-2 and Bcl-x_L as well as 3 therapeutic agents that are commonly used in myeloma therapy; dexamethasone, carfilzomib, and melphalan. We compared the responses of Bim-deficient lines to the Bim-expressing cell lines MM.1s and RPMI8226. As expected LP-1 and KMS26 were highly insensitive to ABT-737 and dexamethasone (IC50s (μM): ABT-737: 8226 – 0.65, MM.1s – 0.5, KMS26 – 3.0, LP-1 – 4.5; Dexamethasone: 8226 – 1.9, MM.1s – 1.5, KMS26 – could not be calculate due to no response at tested concentrations, LP-1 – 19.7) as we and others have previously demonstrated Bim-dependence of these agents in myeloma through silencing experiments. When treated with carfilzomib, LP-1 and KMS26 exhibited an intermediate sensitivity (IC50s: (nM): 8226 – 7.9, MM.1s – 6.0, KMS26 – 20.3, LP-1 – 16.2). Given previous data generated in our lab indicating proteasome inhibitor-induced cell death is inhibited by over-expression of Bcl-2, Bcl-x_L and Mcl-1 and that the apical caspase activated following proteasome inhibition is caspase-8, the apoptotic response to carfilzomib in KMS26 and LP-1 is most consistent with the activation of Bid. Interestingly, when treated with melphalan, all four cell lines exhibited similar sensitivities (IC50s (μM): 8226 – 65, MM.1s – 19.4, KMS26 – 42, LP-1 – 26.7). Previous studies have demonstrated that alkylating agents can induce PARP1-dependent necrosis in the absence of Bax/Bak. Therefore to determine if the melphalan-induced death was caspase-dependent, we co-treated the cell lines with melphalan and the caspase-inhibitor, qVD-OPH. Upon co-treatment, the apoptotic response in all cell lines was drastically reduced, indicating a caspase-dependent death (IC50s (μM): 8226 – 131, MM.1s – 118, KMS26 – 386, LP-1 – 80.5).

In order to verify that these responses are representative of loss of Bim and not unique to LP-1 and KMS26, we utilized ABT-737 resistant cell lines generated in our lab that express very low levels of Bim (KMS11-ABTR and U266-ABTR). Similar to LP-1 and KMS26, these cell lines displayed an intermediate sensitivity to carfilzomib when compared to control sensitive lines (KMS11-CR2 and U266-CR2). Additionally, both KMS11-ABTR and U266-ABTR were sensitive to melphalan in a caspase-dependent manner. We are currently determining the role of Puma and Bid as BH3-activating proteins in proteasome inhibitor- and melphalan-induced cell death.

Our data indicate that active therapeutic agents in myeloma initiate the intrinsic apoptosis pathway through different BH3-activator proteins. Defining the activator proteins for therapeutics could allow for more precise therapeutic decision-making based on patient expression data. Additionally these findings suggest that rational combination therapies could be designed based on combining Bim-dependent and -independent therapies without having to understand how the specific targets of therapies interact (e.g. steroids and alkylating agents).