Bone marrow microenvironment-derived signals induce Mcl-1 dependence in multiple myeloma

Abstract

Multiple myeloma is highly dependent on the bone marrow microenvironment until progressing to very advanced extramedullary stages of the disease such as plasma cell leukemia. Stromal cells in the bone marrow secrete a variety of cytokines that promote plasma cell survival by regulating antiapoptotic members of the Bcl-2 family including Mcl-1, Bcl-x_L, and Bcl-2. Although the antiapoptotic protein on which a cell depends is typically consistent among normal cells of a particular phenotype, Bcl-2 family dependence is highly heterogeneous in multiple myeloma. Although normal plasma cells and most multiple myeloma cells require Mcl-1 for survival, a subset of myeloma is codependent on Bcl-2 and/or Bcl-x_L We investigated the role of the bone marrow microenvironment in determining Bcl-2 family dependence in multiple myeloma. We used the Bcl-2/Bcl-x_L inhibitor ABT-737 to study the factors regulating whether myeloma is Mcl-1 dependent, and thus resistant to ABT-737-induced apoptosis, or Bcl-2/Bcl-x_L codependent, and thus sensitive to ABT-737. We demonstrate that bone marrow stroma is capable of inducing Mcl-1 dependence through the production of the plasma cell survival cytokine interleukin-6 (IL-6). IL-6 upregulates Mcl-1 transcription in a STAT3-dependent manner, although this occurred in a minority of the cells tested. In all cells, IL-6 treatment results in posttranslational modification of the proapoptotic protein Bim. Phosphorylation of Bim shifts its binding from Bcl-2 and Bcl-x_L to Mcl-1, an effect reversed by MEK inhibition. Blocking IL-6 or downstream signaling restored Bcl-2/Bcl-x_L dependence and may therefore represent a clinically useful strategy to enhance the activity of Bcl-2 inhibitors.

Figure 1.

Stromal cells decrease Bcl-2/Bcl-x_Ldependence of multiple myeloma. (A) Myeloma patient bone marrow aspirates were divided into either a buffy coat fraction containing myeloma and stromal cells or CD138⁺ purified myeloma cells and then treated with increasing concentrations of ABT-737 for 24 hours. Apoptosis of myeloma cells was measured by flow cytometry following staining for annexin V-FITC to calculate the IC₅₀ under the 2 conditions. Plasma cells were identified by staining with CD38 and CD45 (CD38⁺, CD45⁻). The IC₅₀ of myeloma in the presence of stromal cells is plotted along the y-axis, and the IC₅₀ of CD138⁺ purified cells is plotted on the x-axis. Each dot represents a single patient sample. Dashed lines represent the cutoff for ABT-737 sensitivity at 0.5 μM. (B) Hs-5 stromal cells and conditioned media induce resistance to ABT-737 but not bortezomib or arsenic trioxide in myeloma cell lines. MM.1s, 8226, KMS18, OPM2, and U266 were treated with the indicated drugs and concentrations for 24 hours in the presence or absence of Hs-5 cells or conditioned media (50%) before staining with CD38 to identify plasma cells and annexin V to measure apoptosis. (C) KMS21BM and OCI-My5 were treated with ABT-199 at the indicated concentrations for 24 hours in the presence or absence of Hs-5 cells or conditioned media (50%) before staining with CD38 to identify plasma cells and annexin V to measure apoptosis. Data are presented as the mean ± standard error (SE) of 3 independent experiments (*P < .05; **P < .01; ***P < .001).

Figure 2.

Stroma-derived IL-6 reduces Bcl-2/Bcl-x_Ldependence. (A-C) MM.1s was treated with the indicated concentrations of ABT-737 for 24 hours, and myeloma cell death determined by staining with annexin V-FITC and CD38. Data are presented as the mean ± SE of 3 independent experiments (*P < .05; **P < .01; ***P < .001; ****P < .0001). (A) Cells were incubated in the presence or absence of Hs-5 cells, Hs-5 conditioned media (CM, 50%), conditioned media from the stromal cells of a myeloma patient (BMSC CM, 50%), or IL-6 (10 ng/mL). (B) Cells were incubated in the presence or absence of Hs-5 conditioned media (0.5%) and the indicated concentrations of neutralizing IL-6 antibody. (C) Cells were incubated in the presence or absence of Hs-5 cells or conditioned media (0.5% or 50%) and the neutralizing anti-IL-6 antibody and/or anti-IL-6 receptor blocking antibody. (D) MM.1s was incubated with BMSCs from a normal donor (NBMSC) in the presence or absence of a neutralizing anti-IL-6 antibody (30 μg/mL) and treated with the indicated concentrations of ABT-737 for 24 hours. Myeloma cell death was determined by staining with annexin V-FITC and CD38. (E) CD138⁺ plasma cells (IPC) purified from a myeloma patient bone marrow aspirate were treated with ABT-737 and IL-6 (10 ng/mL). The total pool of mononuclear cells isolated by Ficoll separation (BCPC) were also treated with ABT-737 and a neutralizing anti-IL-6 antibody (30 μg/mL). Cell viability was determined by flow cytometry after staining with annexin V-FITC. Plasma cells were identified by staining with CD38 and CD45 (CD38⁺, CD45⁻).

Figure 3.

IL-6 regulates Bcl-2 family expression and binding to Bim. MM.1s (A) and KMS18 (B) cells were cocultured with Hs-5 cells, 50% conditioned media from Hs-5 cells, or treated with 10 ng/mL IL-6 for 24 hours. Lysates were prepared and subject to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) followed by western blotting for Bim, Mcl-1, Bcl-x_L, Bcl-2, and actin. (C) The 8226, KMS18, MM.1s, OPM2, and U266 cells were treated with 10 ng/mL IL-6 for 24 hours. Cells were then lysed for protein and western blotting or RNA and quantitative reverse transcription PCR (qRT-PCR). A representative western blot of 3 independent experiments is shown. (D) Changes in protein expression were quantitated by densitometry after normalizing to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) levels. (E) RNA expression of Mcl-1 and Bcl-2 were measured using qRT-PCR. Both graphs depict the mean percent change ± SE relative to untreated cells (*P < .05 by the 1 sample Student t test). (F) MM.1s and KMS18 were treated with 1 ng/mL IL-6 for 24 hours, and protein lysates were prepared then subjected to coimmunoprecipitation with anti-Mcl-1, anti-Bcl-x_L, and anti-Bcl-2 antibodies. The resulting protein complexes and protein input (WCL) were examined by western blot analysis using anti-Bim, anti-Mcl-1, anti-Bcl-x_L, and anti-Bcl-2. The proportion of Bim bound to Mcl-1, Bcl-x_L, and Bcl-2 was quantitated by densitometry and is represented as a pie chart for each condition. Each pie chart is the average of 3 independent experiments.

Figure 4.

STAT3 knockdown partially reverses the protective effect of IL-6. RNA (A) and protein lysates (B) were prepared from cells 72 hours after nucleoporation with either control siRNA or STAT3 siRNA and after 24 hours IL-6 treatment. RNA expression of Mcl-1 and SOCS3 was determined by qRT-PCR and is graphed as fold change relative to untreated control siRNA cells. Protein was subjected to SDS-PAGE and western blotting for phospho-STAT3, total STAT3, Mcl-1, and actin. (C) Forty-eight hours after siRNA nucleoporation, cells were treated with 1.5 μM of ABT-737 with or without 1 ng/mL IL-6 for 24 hours and cell death measured by annexin V-FITC and PI staining. Data are presented as the mean ± SE of 3 independent experiments (**P < .01; ****P < .0001).

Figure 5.

Distinct effects of IL-6 signaling pathways on Bim and Mcl-1. (A) MM.1s and KMS18 were treated with ABT-737 in combination with 1 ng/mL IL-6 and either 1 μM ruxolitinib, 10 μM U0126, or 10 μM LY294002 for 24 hours. Cell death was measured by flow cytometry after staining with annexin V-FITC and PI. Data are presented as the mean ± SE of 3 independent experiments (*P < .05; **P < .01; ***P < .001; ****P < .0001). (B) MM.1s was treated with 1 ng/mL IL-6 in the presence of 1 μM AZD1480, 10 μM LY294002, 10 μM SB203580, 10 μM SP600125, or 10 μM U0126 for 15 minutes. Protein lysates were prepared and subjected to SDS-PAGE and western blotting with antibodies against serine 69 phosphorylated Bim, total Bim, phosphorylated Erk, and total Erk. (C) KMS18 was treated with 1 ng/mL IL-6 for the indicated times and with 10 μM U0126 followed by western blotting for Mcl-1, Bcl-2, Bcl-x_L, serine 69 phosphorylated Bim, total Bim, and actin. (D) KMS18 was treated with 1 ng/mL IL-6 with or without 1 μM ruxolitinb for 24 hours followed by western blotting for Mcl-1, Bcl-2, Bcl-x_L, and GAPDH.

Figure 6.

MEK inhibition increases Bcl-2 dependence. (A) KMS18, (B) MM.1s, (C) KMS12BM, and (D) OCI-My5 were treated with 1 ng/ml IL-6 and/or 10 μM U0126 for 24 hours, and protein lysates were prepared and then subjected to coimmunoprecipitation with anti-Bim antibodies. The resulting protein complexes and protein input (WCL) were examined by western blot analysis using anti-Bim, anti-Mcl-1, anti-Bcl-xL, anti-Bcl-2, and anti-actin. The amount of coprecipitated protein was quantitated by densitometry and normalized to control. Figures are representative of 2 to 3 independent experiments.

Figure 7.

Model of stroma-induced Mcl-1 dependence. See “Discussion” for details.