Co-inhibition of BCL-W and BCL2 restores antiestrogen sensitivity through BECN1 and promotes an autophagy-associated necrosis

Abstract

BCL2 family members affect cell fate decisions in breast cancer but the role of BCL-W (BCL2L2) is unknown. We now show the integrated roles of the antiapoptotic BCL-W and BCL2 in affecting responsiveness to the antiestrogen ICI 182,780 (ICI; Fulvestrant Faslodex), using both molecular (siRNA; shRNA) and pharmacologic (YC137) approaches in three breast cancer variants; MCF-7/LCC1 (ICI sensitive), MCF-7/LCC9 (ICI resistant), and LY2 (ICI resistant). YC137 inhibits BCL-W and BCL2 and restores ICI sensitivity in resistant cells. Co-inhibition of BCL-W and BCL2 is both necessary and sufficient to restore sensitivity to ICI, and explains mechanistically the action of YC137. These data implicate functional cooperation and/or redundancy in signaling between BCL-W and BCL2, and suggest that broad BCL2 family member inhibitors will have greater therapeutic value than targeting only individual proteins. Whereas ICI sensitive MCF-7/LCC1 cells undergo increased apoptosis in response to ICI following BCL-W+/-BCL2 co-inhibition, the consequent resensitization of resistant MCF-7/LCC9 and LY2 cells reflects increases in autophagy (LC3 cleavage; p62/SQSTM1 expression) and necrosis but not apoptosis or cell cycle arrest. Thus, de novo sensitive cells and resensitized resistant cells die through different mechanisms. Following BCL-W+BCL2 co-inhibition, suppression of functional autophagy by 3-methyladenine or BECN1 shRNA reduces ICI-induced necrosis but restores the ability of resistant cells to die through apoptosis. These data demonstrate the plasticity of cell fate mechanisms in breast cancer cells in the context of antiestrogen responsiveness. Restoration of ICI sensitivity in resistant cells appears to occur through an increase in autophagy-associated necrosis. BCL-W, BCL2, and BECN1 integrate important functions in determining antiestrogen responsiveness, and the presence of functional autophagy may influence the balance between apoptosis and necrosis.

Figure 1. Increased expression of BCL-W and BCL2 in MCF-7/LCC9 cells.

Whole cell lysates were subjected to Western blot analysis with a specific BCL2 or BCL-W antibody. (A) Bars represent the mean±SE of the relative BCL2∶actin ratio (normalized to control cells) for three independent experiments. Inset, a representative blot. (B) Bars represent the mean±SE of the relative BCL-W∶actin ratio (normalized to control cells) for three independent experiments. Inset, a representative blot.

Figure 2. BCL-W and BCL2 inhibition increases sensitivity to ICI 182,780 and increases necrosis in MCF-7/LCC9 cells.

(A) Cells were treated with YC137 and/or ICI for 7-days. Bars represent the mean±SE of relative cell proliferation (normalized to EtOH treated controls) for a single representative experiment performed in triplicate. (B) Cells were treated and stained with propidium iodide (PI). Bars represent the mean±SE of relative PI staining (normalized to control EtOH treated cells) for three independent experiments. (C) Cells were transfected with siRNA and stained with PI. Inset, a representative blot showing BCL-W and BCL2 siRNA knockdown.

Figure 3. LC3II and p62/SQSTM1 expression after BCL-W and BCL2 inhibition.

Whole cell lysates were subjected to Western blot analysis with a specific LC3 or p62/SQSTM1 antibody. (A) Bars represent the mean±SE of the relative LC3II∶actin ratio (normalized to empty vector controls) for three independent experiments. Inset, a representative blot. (B) Bars represent the mean±SE of the relative p62/SQSTM1∶actin ratio (normalized to empty vector controls) for three independent experiments. Inset, a representative blot. (C) Cells were transfected with siRNA and LC3II measured by Western blot analysis. Bars represent the mean±SE of the relative LC3II∶actin ratio (normalized to empty vector controls) for three independent experiments.

Figure 4. Increased apoptosis and decreased necrosis after BCL-W and BCL2 and autophagy inhibition.

(A) Cells were treated with ICI, 3MA, YC137, or a combination of the three for 7-days. Bars represent the mean±SE of relative proliferation (normalized to empty vector control). (B) Cells were treated with YC137, ICI, 3MA, YC137+ICI, or a combination of YC137, ICI, and 3MA for prior to Annexin V staining. Bars represent the mean±SE of the relative Annexin V staining (normalized to empty vector controls) for three independent experiments. (C) Cells were treated with ICI, 3MA, YC137, or a combination prior to JC-1 staining. (D) Cells were treated with YC137, ICI, 3MA, YC137+ICI, or a combination of YC137, ICI, and 3MA prior to PI staining. Bars represent the mean±SE of the relative PI staining (normalized to empty vector controls) for three independent experiments.

Figure 5. BCL-W/BCL2 knockdown and autophagy inhibition increases apoptosis and decreases necrosis.

(A) MCF-7/LCC9 cells were transfected with a combination of BCL-W and BCL2 siRNA and treated with ICI and 3MA. Bars represent the mean±SE of the relative Annexin V staining (normalized to empty vector controls) for three independent experiments. (B) Bars represent the mean±SE of the relative PI staining (normalized to empty vector controls) for three independent experiments.

Figure 6. BECN1 knockdown and BCL-W/BCL2 co-inhibition decreases cell proliferation through increased apoptosis in resistant cells.

(A) Whole cell lysates were subjected to Western blot analysis with a specific BECN1 monoclonal antibody. Bars represent the mean±SE of the relative BECN1∶actin ratio (normalized to control cells) for three independent experiments. (B) shRNA infected MCF-7/LCC9 cells were treated with YC137 and/or ICI for 7-days. Bars represent the mean±SE of relative cell proliferation (normalized to EtOH treated controls) for a single representative experiment performed in triplicate. (C) shRNA infected MCF-7/LCC9 cells were treated with YC137, ICI, or a combination of YC137 and ICI. Bars represent the mean±SE of the relative Annexin V staining (normalized to empty vector controls) for three independent experiments. (D) Bars represent the mean±SE of the relative PI staining (normalized to empty vector controls) for three independent experiments.