Metformin induces both caspase-dependent and poly(ADP-ribose) polymerase-dependent cell death in breast cancer cells

Abstract

There is substantial evidence that metformin, a drug used to treat type 2 diabetics, is potentially useful as a therapeutic agent for cancer. However, a better understanding of the molecular mechanisms through which metformin promotes cell-cycle arrest and cell death of cancer cells is necessary. It will also be important to understand how the response of tumor cells differs from normal cells and why some tumor cells are resistant to the effects of metformin. We have found that exposure to metformin induces cell death in all but one line, MDA-MB-231, in a panel of breast cancer cell lines. MCF10A nontransformed breast epithelial cells were resistant to the cytotoxic effects of metformin, even after extended exposure to the drug. In sensitive lines, cell death was mediated by both apoptosis and a caspase-independent mechanism. The caspase-independent pathway involves activation of poly(ADP-ribose) polymerase (PARP) and correlates with enhanced synthesis of PARP and nuclear translocation of apoptosis-inducing factor (AIF), which plays an important role in mediating cell death. Metformin-induced, PARP-dependent cell death is associated with a striking enlargement of mitochondria. Mitochondrial enlargement was observed in all sensitive breast cancer cell lines but not in nontransformed cells or resistant MDA-MB-231. Mitochondrial enlargement was prevented by inhibiting PARP activity or expression. A caspase inhibitor blocked metformin-induced apoptosis but did not affect PARP-dependent cell death or mitochondrial enlargement. Thus, metformin has cytotoxic effects on breast cancer cells through 2 independent pathways. These findings will be pertinent to efforts directed at using metformin or related compounds for cancer therapy.

Figures

Figure 1. Metformin induces cell death in breast cancer cells

A. MCF7 cells were treated with metformin (8mM) for the indicated time. Live cells were counted using a trypan blue exclusion assay to obtain a growth curve for control (solid line) and metformin treated cells (dashed line). B. MCF7 and T47D were treated without (−) or with 8 mM metformin (+) for 3 days and MDA-MB-453, BT474, MDA-MB-231 and MCF10A were treated for 4 days. The percent of dead cells in the population was determined by trypan blue exclusion.

Figure 2. Cell death caused by metformin in MCF7 cells has features of apoptosis

A. MCF7 cells were treated with (+) or without (−) metformin for 1 (D1), 2 (D2), or 3 (D3) days and then western blotting was performed to detect cleaved PARP1, cleaved caspase 7, and β-actin. B. Left panel: MCF7 cells were treated with (Met) or without (Con) metformin for 3 days and cells were stained with PI and Annexin V-FITC (top panels). Cells were visualized by fluorescence microscopy using a dual-filter set that allows simultaneous detection of both fluorescent signals. Photographs were taken at 100 X magnification. Green indicates Annexin V positive cells and red indicates PI positive cells. Bottom panels show phase contrast images of the same cultures. Right panel: MCF7 cells were treated with metformin (8 mM) for three days and cells were stained with PI and Annexin V-FITC and analyzed by flow cytometry. Percentage of Annexin V positive cells is shown. C. MCF7 cells were treated with (+) or without (−) metformin (8mM) in the presence or absence (DMSO) or a pan-caspase inhibitor (Q-Val-Asp-OPh, 10 µM) for 2.5 days. Western blotting was performed to examine cleaved PARP1 and caspase 7. β-Actin was detected as a loading control. D. MCF7 cells were treated as in C for either 2.5 days (left panel) or 4 days (right panel). The percentage of dead cells in each culture was obtained using the trypan blue exclusion assay.

Figure 3. Metformin causes enlargement of mitochondria in breast cancer cells that are sensitive to metformin treatment

A. MCF7 and T47D were treated with (Met) or without (Con) metformin for 3 days and MDA-MB-453, BT474, MDA-MB-231, and MCF10A were treated for 4 days. Photographs are phase contrast images taken at 200X magnification. B. MCF7 cells were treated as in A and cells were fixed and processed for transmission electron microscopy. Photographs were taken at the magnification indicated in each panel. C. MCF7 cells were electroporated with a plasmid encoding DsRed-Mito. Cells were treated with (Met) or without (Con) metformin for 2.5 days. Cells were fixed and observed under confocal microscopy. Both panels were photographed at the same magnification.

Figure 4. Metformin induces PARP-dependent cell death in MCF7 cells

A. Left panel: MCF7 cells were treated for 4 days with (+) or without (−) metformin in the presence of either vehicle (DMSO) or PARP inhibitor (PARP Inh). Right panel: MCF7 cells were treated for 4 days with (+) or without (−) metformin in the presence of a pan-caspase inhibitor (Q-Val-Asp-OPh, 10 µM) and either vehicle (DMSO) or PARP inhibitor (PARP Inh). The percentage of dead cells in each dish was obtained via trypan blue exclusion assay. A t-test was used to determine statistical differences between the indicated groups (* indicates a P value < 0.02; ** indicates a P value < 0.001). B. MCF7 cells stably expressing either a control shRNA (shLuc) or an shRNA targeting PARP1 (shPARP) were treated with metformin and pan-caspase inhibitor as indicated for either 2 days (left panel) or 3 days (right panel). The percentage of dead cells in each dish was obtained via trypan blue exclusion assay. C. Extracts from cells treated as in B were analyzed by western blotting for PARP1, cleaved caspase 7, full-length caspase 7, and β-actin.

Figure 5. Metformin-induced mitochondrial enlargement is associated with PARP-dependent cell death

A. MCF7 cells were treated with or without (Control) metformin for 2.5 days in the presence of vehicle (No Inhibitor), pan-caspase inhibitor, PARP inhibitor, or both pan-caspase inhibitor and PARP inhibitor. Phase contrast photomicrographs were taken at 200X magnification. B. MCF7 cells expressing pAcGFP1-Mito were treated as in A and observed by confocal microscopy. C. MCF7 cells stably expressing either a control shRNA (shLuc) or an shRNA targeting PARP1 (shPARP) were treated for 2 days with or without metformin in the presence of DMSO (Vehicle) or caspase inhibitor. Phase contrast photomicrographs were taken as in A.

Figure 6. Metformin stimulates PAR synthesis in sensitive cells

A. MCF7 (left panel) or MDA-MB-231 (right panel) cells were treated with (Met) or without (Con) metformin for 2.5 days and then PAR (red) was detected by immunofluorescence using confocal microscopy. Nuclei (blue) were stained with DAPI. B. MCF7 cells were treated with or without metformin in the presence of DMSO or PARP inhibitor for 2.5 days. PAR (red) and nuclei (blue) were detected as described in A.

Figure 7. AIF plays a role in mediating sensitivity to metformin and the drug promotes AIF release from mitochondria and uptake into nuclei

A. MCF7 cells were treated with or without metformin in the presence or absence of PARP inhibitor for 2.5 days. Cells were then fixed and immunofluorescence was used to detect endogenous AIF (green signal). B. MCF7 cells were treated with (Met) or without (Con) metformin for 2.5 days and then nuclear levels of AIF were examined by western blotting. Sp1, a transcription factor, was used as a nuclear marker and loading control. C. MCF7 cells were transfected with an siRNA that targets AIF (siAIF) or a non-targeted control siRNA (Scramble). One day after transfection the cells were treated with (M) or without metformin (C) and treated with vehicle (DMSO) or pan-caspase inhibitor (Casp Inh). After 3 days the cells were harvested and analyzed by western blotting to detect AIF or β-actin. D. The number of live and dead cells in cultures transfected and treated as described in B was determined by trypan blue exclusion assays. A t-test was used to determine statistical differences between the indicated groups (* indicates a P value < 0.05).