Mitochondrial VDAC1 Silencing Leads to Metabolic Rewiring and the Reprogramming of Tumour Cells into Advanced Differentiated States

Abstract

Oncogenic properties, along with the metabolic reprogramming necessary for tumour growth and motility, are acquired by cancer cells. Thus, tumour metabolism is becoming a target for cancer therapy. Here, cancer cell metabolism was tackled by silencing the expression of voltage-dependent anion channel 1 (VDAC1), a mitochondrial protein that controls cell energy, as well as metabolic and survival pathways and that is often over-expressed in many cancers. We demonstrated that silencing VDAC1 expression using human-specific siRNA (si-hVDAC1) inhibited cancer cell growth, both in vitro and in mouse xenograft models of human glioblastoma (U-87MG), lung cancer (A549), and triple negative breast cancer (MDA-MB-231). Importantly, treatment with si-hVDAC1 induced metabolic rewiring of the cancer cells, reversing their oncogenic properties and diverting them towards differentiated-like cells. The si-hVDAC1-treated residual "tumour" showed reprogrammed metabolism, decreased proliferation, inhibited stemness and altered expression of genes and proteins, leading to cell differentiation toward less malignant lineages. These VDAC1 depletion-mediated effects involved alterations in master transcription factors associated with cancer hallmarks, such as highly increased expression of p53 and decreased expression of HIF-1a and c-Myc that regulate signalling pathways (e.g., AMPK, mTOR). High expression of p53 and the pro-apoptotic proteins cytochrome c and caspases without induction of apoptosis points to functions for these proteins in promoting cell differentiation. These results clearly show that VDAC1 depletion similarly leads to a rewiring of cancer cell metabolism in breast and lung cancer and glioblastoma, regardless of origin or mutational status. This metabolic reprogramming results in cell growth arrest and inhibited tumour growth while encouraging cell differentiation, thus generating cells with decreased proliferation capacity. These results further suggest VDAC1 to be an innovative and markedly potent therapeutic target.

Keywords: VDAC1; cancer stem cells; differentiation; mitochondria; si-RNA.

Figure 1

si-hVDAC1 treatment silences VDAC1 expression, causes cell growth inhibition and reduces energy production. (A) IHC staining of VDAC1 in sections derived from human normal tissue (n = 13), glioblastoma (n = 40), lung cancer (n = 20) and breast cancer (n = 20) in tissue microarray slides (Biomax). Percentages of sections stained at the indicated intensity are shown. (B, C) U-87MG, A549 and MDA-MB-231 cells were treated with 50 nM si-NT (black bars) or si-hVDAC1 (grey bars) and 72 h post-treatment were analysed for VDAC1 levels by immunoblotting (B) and cell growth using the SRB assay (mean ± SEM; n = 3) (C). (D, E) WI-38 and HaCaT cells treated with si-NT (50 or 75 nM, black and grey bars, respectively) or si-hVDAC1 (50 or 75 nM, light grey and white bars, respectively) and analysed for VDAC1 levels by immunoblotting 48 h post-transfection (RU indicates relative value) (D) and for cell growth using the SRB assay (mean ± SEM; n = 3) (E). F, G) U-87MG (black bars), A549 (light grey bars) and MDA-MB-231 cells (white bars) were transfected with si-NT or si-hVDAC1 (50 nM) and 24 h post-transfection, the cells were again transfected with plasmid pcDNA4/TO, either empty or encoding mVDAC1. After 24 h, cell growth was analysed by the SRB method (mean ± SEM; n = 3) (F) or analysed for VDAC1 levels by immunoblotting (G). (H–J) Immunoblot (H), mitochondrial membrane potential (ΔΨ) (I) and ATP (J) levels were analysed in U-87MG, A549 and MDA-MB-231 cells treated with 50 nM si-NT (black bars) or si-hVDAC1 (grey bars). Cells treated with FCCP, (25 μM) (white bars) served as controls for decreasing ΔΨ and ATP levels. β-actin served as an internal loading control. Mean ± SEM; n = 3; * p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001.

Figure 2

si-hVDAC1 inhibits GBM-, A549 lung cancer- and MDA-MB-231 breast cancer-derived tumour growth in a xenograft mouse model. (A–C) U-87MG (A), A549 (B) and MDA-MB-231 (C) cells were subcutaneously inoculated into nude mice. When tumour size reached 50-100 mm³, the mice were divided into 2 matched groups and xenografts were injected intratumourally every 3 days with si-NT (•, 4–5 mice) or si-hVDAC1 (○, 3–6 mice) to a final concentration of 50–60 nM. The calculated average tumour volumes are presented as means ± SEM. (D, E) si-NT-TT and si-hVDAC1-TT sections from U-87MG, A549 and MDA-MB-231 xenograft mice were stained for VDAC1 by IHC (D) or subjected to immunoblot (E). RU = average relative units, presented as the mean ± SEM; n = 3–4 mice. β-actin served as an internal loading control. (F) Levels of VDAC1, VDAC2 and VDAC3 mRNA in si-hVDAC1-TTs from U-87MG, A549 and MDA-MB-231 cells, as analysed by q-RT-PCR and presented relative to the levels seen in si-NT-TTs. (G) Representative IHC staining of si-NT-TTs and si-hVDAC1-TTs derived from U-87MG-, A549- and MDA-MB-231 cells with anti-Ki-67 antibodies. (H, I) Quantitative analysis of Ki-67-positive cells (H) and Ki-67 mRNA levels (I) in U-87MG-, A549- and MDA-MB-231-derived tumours presented as fold of decrease. Results show the mean ± SEM (n = 3–5), p: * ≤ 0.05; ** ≤ 0.01, *** ≤ 0.001.

Figure 3

si-hVDAC1 treatment reverses the reprogrammed metabolism of U-87MG-, A549- and MDA-MB-231-derived tumours. (A–C) Representative IHC staining using specific antibodies against Glut-1, GAPDH, CS, complex IVc (comp. IVc) and ATP synthase 5a (ATP syn. 5a) from si-NT-TTs or si-hVDAC1-TTs sections derived from U-87MG (A), A549 (B) and MDA-MB-231 (C) xenografts. (D–F) Immunoblots and quantitative analysis of selected metabolism-related proteins from si-NT-TT or si-hVDAC1-TT sections from U-87MG (D), A549 (E) and MDA-MB-231 (F) xenografts. Quantitative analysis is presented as the mean ± SEM in relative units (RU), (n = 3–4 mice). β-actin served as an internal loading control. (G–I) mRNA levels of metabolic enzymes in si-hVDAC1-TTs, relative to those in si-NT-TTs derived from U-87MG (G), A549 (H) and MDA-MB-231 (I) tumours, represented as fold change. Results are means ± SEM (n = 5 tumours for each), p: * ≤ 0.05; ** ≤ 0.001; *** ≤ 0.0001. (J–L) Immunoblots and quantitative analysis of phosphorylated AMPK, phospho-S6 (pS6) and SIRT1 using specific antibodies in si-hVDAC1-TTs and si-NT-TTs derived from U-87MG (J), A549 (K) and MDA-MB-231 (L) cells. RU, relative units; the mean ± SEM (n = 3–4 mice) are shown, p: * ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001, **** ≤ 0.0001; β-actin as an internal loading control are shown.

Figure 3

si-hVDAC1 treatment reverses the reprogrammed metabolism of U-87MG-, A549- and MDA-MB-231-derived tumours. (A–C) Representative IHC staining using specific antibodies against Glut-1, GAPDH, CS, complex IVc (comp. IVc) and ATP synthase 5a (ATP syn. 5a) from si-NT-TTs or si-hVDAC1-TTs sections derived from U-87MG (A), A549 (B) and MDA-MB-231 (C) xenografts. (D–F) Immunoblots and quantitative analysis of selected metabolism-related proteins from si-NT-TT or si-hVDAC1-TT sections from U-87MG (D), A549 (E) and MDA-MB-231 (F) xenografts. Quantitative analysis is presented as the mean ± SEM in relative units (RU), (n = 3–4 mice). β-actin served as an internal loading control. (G–I) mRNA levels of metabolic enzymes in si-hVDAC1-TTs, relative to those in si-NT-TTs derived from U-87MG (G), A549 (H) and MDA-MB-231 (I) tumours, represented as fold change. Results are means ± SEM (n = 5 tumours for each), p: * ≤ 0.05; ** ≤ 0.001; *** ≤ 0.0001. (J–L) Immunoblots and quantitative analysis of phosphorylated AMPK, phospho-S6 (pS6) and SIRT1 using specific antibodies in si-hVDAC1-TTs and si-NT-TTs derived from U-87MG (J), A549 (K) and MDA-MB-231 (L) cells. RU, relative units; the mean ± SEM (n = 3–4 mice) are shown, p: * ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001, **** ≤ 0.0001; β-actin as an internal loading control are shown.

Figure 4

si-hVDAC1 treatment markedly reduces cancer stem cell marker expression in U-87MG-, A549- and MDA-MB-231-derived tumours. Representative IHC staining with cell line-specific CSC markers using specific antibodies in si-NT-TT or si-hVDAC1-TT sections derived from U-87MG (A), A549 (E,) and MDA-MB-231 (I) xenografts. Immunoblot of protein extracts obtained from si-NT-TTs or si-hVDAC1-TTs derived from U-87MG (B, C), A549 (F, G) and MDA-MB-231 (J, K) xenografts, using the specific antibodies indicated, and their quantitative analysis presented as relative units (RU); Results are the mean ± SEM (n = 3–4 mice) are shown. β-actin served as an internal loading control. mRNA levels of the indicated genes in si-hVDAC1-TTs relative to those in si-NT-TTs derived from U-87MG (D), A549 (H) and MDA-MB-231 (L) tumours. Results are means ± SEM (n = 3–5 tumours), p: * ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001. Dashed lines indicate the control level.

Figure 5

si-hVDAC1 treatment of U-87MG-, A549- and MDA-MB-231-derived tumours induces expression of proteins associated with differentiation. U-87MG cells-derived hVDAC1-TTs or si-NT-TTs were stained with anti-TUBB3 (A) anti-GAD-67 (B) or anti-GFAP antibodies (C) with representative images from 2 mice each shown. A549 cell-derived hVDAC1-TTs or si-NT-TTs were stained for surfactant C (D) or subjected to q-RT-PCR to analyse expression levels of the indicated genes (E). MDA-MB-231 cell-derived hVDAC1-TTs or si-NT-TTs were stained with anti-CD44, -CD24 or -ERBB2/Her2 antibodies (G). mRNA levels of the indicated genes were analysed in si-hVDAC1-TTs and are presented relative to those in si-NT-TTs (H). Results are means ± SEM (n = 3); p: * ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001. Dashed line indicates the control level.

Figure 6

si-hVDAC1 treatment alters the expression levels of p53, HIF-1α and c-Myc transcription factors and of phosphorylated NF-κB/RelA, p65 in U-87MG-, A549- and MDA-MB-231-derived tumours. (A) Representative IHC staining of p53 in si-NT-TT or si-hVDAC1-TT sections derived from U-87MG, A549 and MDA-MB-231 xenografts. (B–E) Immunoblot analysis of p53, c-Myc and phosphorylated NF-κB/RelA (p65) in si-NT-TTs and si-hVDAC1-TTs derived from U-87MG (B), A549 (C) and MDA-MB-231 (D) xenografts. Quantitative analysis of p53, c-Myc and p65 levels in U-87MG, A549 and MDA-MB-231 cell tumours is presented (E). (F) q-RT-PCR analysis of p53, HIF-1α and c-Myc mRNA levels in U-87MG-, A549- and MDA-MB-231-derived tumours treated with si-hVDAC1, relative to those treated with si-NT. (G) q-RT-PCR analysis of p21 and MDM2 mRNA levels in U-87MG-, A549- and MDA-MB-231-derived tumours treated with si-hVDAC1, relative to those treated with si-NT. (H,I) Immunoblot staining (H) of γ-H2AX (phospho S139) using specific antibodies and their quantitative analysis (I) presented as relative units (RU); mean ± SEM (n = 3 mice) are shown for si-NT-TTs or si-hVDAC1-TTs derived from U-87MG, A549 and MDA-MB-231 xenografts. Results are means ± SEM); p: * ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001. Dashed line indicates the control level.

Figure 7

si-hVDAC1 treatment alters the expression levels of SMAC, Bcl-xL, BAX, caspase 8, caspase 3 and cytochrome c in U-87MG-, A549- and MDA-MB-231-derived tumours. (A) Representative sections of TUNEL staining of si-NT-TT or si-hVDAC1-TT sections derived from U-87MG, A549 and MDA-MB-231 xenografts. Immunoblots (B–D) of SMAC, Bcl-xL, caspase 8, caspase 3 and BAX and their quantitative analysis (E–G) in si-NT-TTs and si-hVDAC1-TTs derived from U-87MG (B, E), A549 (C, F) and MDA-MB-231 (D, G) xenografts. Average relative units (RUs) and β-actin as an internal loading control are shown. Results are means ± SEM (n = 3); p: ** ≤ 0.01; *** ≤ 0.001; **** ≤ 0.0001. (H) Representative cytochrome c staining of si-NT-TT and si-hVDAC1-TT sections derived from U-87MG, A549 and MDA-MB-231 xenografts.

Figure 8

A schematic presentation of cancer cell VDAC1-depletion and metabolic reprogramming leading to a reversal of oncogenic properties and cell differentiation in U-87MG-, A549- and MDA-MB-231-derived tumours. The homeostatic energy and metabolic states of cancer cells in the tested U-87MG, A549 and MDA-MB-231 xenografts are modified upon silencing VDAC1 expression, leading to a reprogramming of metabolism, thereby decreasing energy and metabolite generation (A). This leads to changes in the levels of the master metabolism regulator p53, as well as HIF1-α and c-Myc expression levels, and of other genes (B), resulting in a reversal of oncogenic properties, including elimination of CSCs while leading to cell differentiation (C), thereby preventing tumour recurrence. Tumour treatment with conventional therapy, such as chemotherapy or radiation, targets cancer cells but not cancer stem cells (D). As such, tumour relapse may occur.