Metformin is widely used to reduce the high blood sugar levels caused by diabetes. Recently, several studies have suggested that patients taking metformin who also develop cancer have tumors that grow more slowly than average. As clinical trials have already started to investigate if metformin is an effective anti-cancer treatment, it is important to understand how it might restrict tumor growth.

Researchers have proposed two ways that metformin could affect tumors. First, insulin is known to prompt cancer cells to divide, so the slower rate of tumor growth could just be a side-effect of the metformin reducing the amount of insulin in the blood. Alternatively, metformin could target cancer cells more directly by cutting the energy supply produced by their mitochondria. Metformin has been shown to disrupt complex I of the electron transport chain that is used by cells to generate energy. However, it is not known if disrupting complex I would actually stop cells dividing because they can generate energy in other ways.

Wheaton, Weinberg et al. have now demonstrated that metformin does target complex I in cancer cells, and that its effects depend on the amount of glucose available for cells to convert, without involving mitochondria, into energy. When there is plenty of glucose, metformin slows down the rate at which cancer cells divide, which slows down tumor growth. When the cells are deprived of glucose, metformin kills the cells instead.

Metformin also inhibits the pathways that regulate hypoxia inducible factors (HIFs), which are part of a system that helps cells to survive low-oxygen conditions, a prominent feature of many tumors. This means that metformin may combat cancer more effectively if used alongside other treatments that reduce the availability of both oxygen and glucose inside cells. Metformin could also potentially treat conditions that are linked to overactive HIFs, such as pulmonary hypertension.

Metformin is widely used to treat patients with type II diabetes mellitus who have high levels of circulating insulin (Nathan et al., 2009). Metformin suppresses liver gluconeogenesis thereby reducing glucose release from the liver (Inzucchi et al., 1998; Viollet et al., 2012). In several recent retrospective studies, investigators have observed an association between metformin use and diminished tumor progression in patients suffering from different types of cancers (Evans et al., 2005; Bowker et al., 2006; Dowling et al., 2012). These data have prompted several prospective clinical trials to determine the efficacy of metformin as an anti-cancer agent. However, the underlying mechanism by which metformin diminishes tumor growth is not fully understood.

There are two postulated mechanisms by which metformin reduces tumor growth. Metformin may act at the organismal level, reducing levels of circulating insulin, a known mitogen for cancer cells. Alternatively, metformin may act in a cancer cell autonomous manner. Metformin is known to inhibit mitochondrial complex I in vitro (Ota et al., 2009; El-Mir et al., 2000; Owen et al., 2000) and it is thus possible that this targeting of the electron transport chain could inhibit tumor cell growth (Birsoy et al., 2012). This latter hypothesis has been questioned as cancer cells have the ability to survive on ATP produced exclusively by glycolysis. Furthermore, cancer cells have been shown to conduct glutamine-dependent reductive carboxylation to generate the TCA cycle intermediates required for cell proliferation when the electron transport chain is inhibited (Mullen et al., 2012; Fendt et al., 2013). Thus, it is not clear whether inhibition of complex I by metformin would result in decreasing tumor growth. In the present study, we directly tested whether inhibition of cancer cell mitochondrial complex I by metformin was required to decrease cell proliferation in vitro and tumor progression in vivo.

Human HCT116 p53^−/− colon cancer cells have previously shown to be sensitive to metformin (Buzzai et al., 2007). To determine if metformin treatment inhibited cellular oxygen consumption in these cells, we treated HCT116 p53^−/− cells with increasing concentrations of metformin in media containing the metabolic substrates glucose, pyruvate, and glutamine for 24 hr. Subsequently, we measured cellular oxygen consumption. Metformin inhibited cellular oxygen consumption of HCT 116 p53^−/− cells at concentrations (0.25–1.0 mM) similar to those reported to affect bioenergetics and gluconeogenesis in primary hepatocytes in vitro (Figure 1A; Foretz et al., 2010; Miller et al., 2013).

Figure 1 with 1 supplement see all

Download asset Open asset

To determine whether metformin's inhibition of cellular oxygen consumption depended on mitochondrial complex I, we stably overexpressed the Saccharomyces cerevisiae protein NDI1 in HCT 116 p53^−/− cells (hereon referred to as NDI1-HCT 116 p53^−/− cells). NDI1 is a single-subunit NADH dehydrogenase, which oxidizes NADH in a process similar to the multi-subunit mammalian complex I; however without proton pumping or ROS generation (Seo et al., 1998). By contrast, mammalian complex I contains 45 subunits that pumps protons and generates ROS. NDI1-HCT 116 p53^−/− cells demonstrated a slight, non-significant elevation in basal cellular oxygen consumption compared to control cells and were completely resistant to the effects of metformin on cellular oxygen consumption (Figure 1—figure supplement 1, Figure 1B).

To ensure that the inhibition of cellular oxygen consumption by metformin was a direct effect of metformin on complex I, we examined mitochondrial respiratory function in saponin-permeabilized cells. Saponin removes cholesterol from plasma membranes, allowing the entry of metabolic substrates directly to mitochondria (Jamur and Oliver, 2010). In the presence of ADP and the complex I substrates pyruvate and malate, metformin fully inhibited oxygen consumption in permeabilized Control-HCT 116 p53^−/− cells (Figure 1C). By contrast, metformin had no effect on pyruvate/malate-driven oxygen consumption in NDI1-HCT 116 p53^−/− cells (Figure 1D). Metformin also had no effect on oxygen consumption in saponin-permeabilized cells respiring on the complex II substrate succinate in the presence of ADP (Figure 1E). Interestingly, in saponin-permeabilized cells, metformin significantly inhibited complex I-dependent respiration at a much lower concentration than that required to inhibit oxygen consumption of intact cells, suggesting that transport across the plasma membrane is a barrier to metformin's inhibition of complex I. Metformin is known to slowly accumulate in cells in which its uptake is mediated by organic cation transporters (OCTs) (Emami Riedmaier et al., 2013). To ensure that NDI1-HCT 116 p53^−/− cells are not refractory to metformin because of a change in metformin uptake, we analyzed the expression of OCT 1 in both control and NDI1-HCT 116 p53^−/− cells. Expression of OCT1 protein did not change with the presence of NDI1 (Figure 1F).

We next sought to determine if metformin-dependent inhibition of complex I resulted in changes in proliferation and survival of HCT116 p53^−/− cells. Metformin did not induce cell death in Control-HCT 116 p53^−/− or NDI1-HCT 116 p53^−/− cells in the presence of glucose (Figure 2A,B), however, in the absence of glucose, metformin induced cell death in Control-HCT 116 p53^−/− but not in NDI1-HCT 116 p53^−/− cells (Figure 2C,D). Metformin diminished cell proliferation in Control-HCT 116 p53^−/− cells but not in NDI1-HCT 116 p53^−/− cells in media containing glucose (Figure 2E,F).

Figure 2 with 2 supplements see all

Download asset Open asset

These results indicate that the metformin-dependent inhibition of complex I decreases cell proliferation in the presence of glucose and increases cell death under glucose deprivation. These inhibitory effects of metformin were not specific to HCT116 p53^−/− cells as metformin inhibited oxygen consumption and cellular proliferation of Control-HCT 116 p53^+/+ cells and Control-A549 human lung cancer cells but not NDI1-HCT 116 p53^+/+ or NDI1-A549 cells (Figure 2—figure supplement 1 and 2). Taken together, these results indicate that the anti-proliferative and cell death promoting effects of metformin require mitochondrial complex I inhibition.

We also examined whether phenformin, a more lipophilic biguanide, also exert its anti-proliferative effects on cancer cells through inhibition of complex I. Phenformin inhibited oxygen consumption in Control-HCT 116 p53^−/− cells and saponin-permeabilized Control HCT 116 p53^−/− cells at 100-fold lower concentration compared to metformin (Figure 3A,C). Expression of NDI1 rescued the phenformin-mediated decrease in oxygen consumption (Figure 3B,D). Phenformin diminished cell proliferation in the control but not NDI1 expressing HCT116 p53^−/− cells (Figure 3E,F), and did not induce cell death in media containing glucose, similar to metformin (Figure 3G,H). Collectively, these results indicate that phenformin also exerts its biological effects through inhibition of mitochondrial complex I.

Figure 3 with 2 supplements see all

Download asset Open asset

To determine whether metformin and phenformin diminish proliferation and survival of cells lacking endogenous complex I activity, we utilized a variant of CCL16 hamster fibroblasts that harbors a mutation in complex I (B2-CCL16) (Seo et al., 1998). Metformin inhibited proliferation of wild-type Control-CCL16 hamster fibroblasts but not that of B2-CCL16 cells or of B2-CCL16 cells reconstituted with NDI1 (NDI1-CCL16). Metformin and phenformin inhibited cellular oxygen consumption in wild-type CCL16 but not in CCL16-NDI1 cells (Figure 3—figure supplement 1 and 2). When these cells were cultured in galactose-substituted media, both metformin and phenformin induced cell death only in wild-type CCL16 cells. Survival of NDI1-CCL16 cells was not affected by metformin or phenformin (Figure 3—figure supplements 1 and 2). The B2-CCL16 cells die in galactose in the absence of metformin or phenformin since they harbor a mutation in complex I. Taken together, these results confirm that the anti-proliferative effects of metformin and phenformin require mitochondrial complex I inhibition.

The positive charge of metformin has been proposed to account for it is accumulation within the matrix of mitochondria that exhibit a robust inner mitochondria membrane potential (Owen et al., 2000). Alternatively, the non-polar hydrocarbon-side chain of the drug could promote binding to complexes within mitochondrial membranes. We tested whether the mitochondrial membrane potential is necessary for metformin-dependent inhibition of complex I. Saponin-permeabilized Control-HCT 116 p53^−/− cells were induced to respire on pyruvate/malate in the presence of either ADP or CCCP. Although both ADP and CCCP induce mitochondrial respiration, only CCCP depolarizes mitochondrial inner membrane potential. Metformin inhibited ADP but not CCCP stimulated oxygen consumption indicating that the metformin-mediated inhibition of mitochondrial complex I required polarized mitochondria (Figure 4A,B). Rotenone, an irreversible inhibitor of complex I, does not require polarized mitochondria to inhibit mitochondrial oxygen consumption (Figure 4C,D). Our results suggest that metformin would not be effective in suppressing complex I activity of intact cells if mitochondrial inner membrane potential was disrupted. Metformin-mediated inhibition of the electron transport chain diminishes proton pumping, which might depolarize the mitochondrial membrane, thus limiting accumulation of the drug. However, we did not observe a reduction in the mitochondrial inner membrane potential measured using TMRE fluorescent dye in Control and NDI1 HCT116 p53^−/− cells after metformin treatment (Figure 4E,F). When electron transport function is inhibited, the ATP synthase can function in reverse such that it uses ATP generated by glycolysis to pump protons across the inner mitochondrial membrane, maintaining membrane potential (Appleby et al., 1999). The ATP synthase inhibitor, Oligomycin A, diminished TMRE fluorescence in Control-HCT 116 p53^−/− cells treated with metformin suggesting that in the presence of metformin, intact cells maintain their mitochondrial membrane potential by reversal of the ATP synthase (Figure 4E).

Figure 4

Download asset Open asset

Rotenone irreversibly inhibits complex I, which contributes to its high toxicity in vivo. Because metformin is well tolerated, we sought to determine if metformin might reversibly bind to complex I. Saponin-permeabilized Control-HCT 116 p53^−/− cells were treated with pyruvate and malate to maintain the mitochondrial inner membrane potential. Metformin was then added, followed by injection of either ADP or CCCP. Metformin inhibited ADP, but not CCCP-stimulated oxygen consumption (Figure 5A,B). As metformin accumulation requires mitochondrial membrane polarization (Figure 4A,B), these results indicate that metformin reversibly inhibits mitochondrial complex I. If the metformin that accumulated in the mitochondrial matrix irreversibly inhibited complex I, then oxygen consumption would have remained attenuated in CCCP-treated cells after metformin treatment.

Figure 5

Download asset Open asset

An emerging function of mitochondria distinct from their ability to perform biosynthetic and bioenergetic reactions is the generation of H₂O₂, which promotes signaling in normal and cancer cells (Hamanaka and Chandel, 2010; Sena and Chandel, 2012). Mitochondrial complexes I and III produce superoxide into the mitochondrial matrix, where it is converted quickly to H₂O2 by SOD2 (Brand, 2010). Mitochondrial complex III also generates superoxide into the mitochondrial intermembrane space where it escapes through VDACs to cytosol and is converted into H₂O₂ by SOD1 (Han et al., 2003; Muller et al., 2004). We measured production and subsequent release of H₂O₂ from isolated mitochondria in the presence of metformin, rotenone, or antimycin A (complex III inhibitor) using pyruvate and malate as substrates.

Consistent with previous reports, rotenone and antimycin increased the release of H₂O₂ from mitochondria isolated from Control-HCT 116 p53^−/− cells (Figure 6A,B; St-Pierre et al., 2002; Muller et al., 2004). In contrast, metformin did not substantially increase H₂O₂ release, suggesting that metformin and rotenone act on different sites of complex I (Figure 6A). When mitochondria were isolated from NDI1–HCT 116 p53^−/− cells, only antimycin lead to a significant increase in H₂O₂ release (Figure 6B). Previous reports have shown that metformin does not substantially increase H₂O₂ production in isolated liver mitochondria and that metformin diminishes mitochondrial H₂O₂ production in response to paraquat, which induces mitochondrial ROS production (Batandier et al., 2006; Algire et al., 2012).

Figure 6

Download asset Open asset

One biological consequence of mitochondrial-generated H₂O₂ is hypoxic stabilization of the hypoxia-inducible factors (HIFs) (Chandel et al., 2000). HIFs are involved in metabolic adaptation of tumor cells to hypoxia (Semenza, 2012). Metformin reduced hypoxic stabilization of HIF-1α in Control-HCT 116 p53^−/− but not in NDI1-HCT 116 p53^−/− (Figure 6C,D). Metformin did not decrease deferoxamine (DFO) stabilization of HIF-1α protein. DFO is an iron chelator known to directly stabilize HIF-1α protein independent of upstream signaling events. Metformin also significantly diminished hypoxic activation of HIF-dependent target genes, vascular endothelial growth factor (VEGF), and carbonic anhydrase 9 (CA9) in Control-HCT 116 p53^−/− but not in NDI1-HCT 116 p53^−/− (Figure 6E). Thus, metformin is an effective agent to reduce hypoxic activation of HIF-1.

Finally, we directly tested whether tumor cell autonomous inhibition of mitochondrial complex I by metformin was required to decrease tumor progression in vivo. As our NDI1-HCT 116 p53^−/− cells are refractory to multiple effects of metformin in vitro, we reasoned that if metformin acted directly on mitochondrial complex I within the tumor cells to reduce tumorigenesis then NDI1-HCT 116 p53^−/− xenograft tumors would not be inhibited in their growth. However, if metformin acts at the organismal level to diminish tumorigenesis then NDI1-HCT 116 p53^−/− xenograft tumor growth would be suppressed similar to control tumors. Control-HCT 116 p53^−/− cells subcutaneously injected into the left flank of nude mice rapidly grew in vivo, while tumors from mice fed metformin through drinking water ad libitum starting 4 days post-implantation exhibited a marked reduction in growth (Figure 7A,B). NDI1-HCT 116 p53^−/− xenograft growth was resistant to metformin therapy (Figure 7A,B), suggesting that the metformin carries out its tumor inhibitory effects in a cancer cell autonomous manner through inhibition of mitochondrial complex I. Importantly, the consumption of water containing metformin was similar between control and NDI1 tumor barring mice (Figure 7C). Transcripts for the HIF target genes CA9 and VEGF were diminished in control tumors treated with metformin but not in NDI1 expressing tumors (Figure 7D,E). The blood glucose, plasma lactate, insulin, and IGF-1 level displayed no differences between the metformin-treated animals and control animals at the end of the study (Figure 7—figure supplement 1), consistent with previous reports (Tomimoto et al., 2008).

Figure 7 with 3 supplements see all

Download asset Open asset

To further bolster our conclusions, we examined tumor growth of A549 cells expressing NDI1 and shRNA targeting mammalian NDUFS3, a subunit of the human mitochondrial complex I (Vogel et al., 2007). A549 cells are null for the tumor suppressor LKB1 and are known to be responsive to metformin therapy (Rocha et al., 2011). Furthermore, LKB1-deficient tumors are more susceptible to the related biguanide phenformin (Shackelford et al., 2013). We replaced the endogenous mitochondrial complex I by expressing the NDI1 protein in A549 cells stably expressing shRNA against NDUFS3 (Figure 7—figure supplement 2A), referred to as NDI1-shNDUFS3-A549 cells. The Control-A549 cells contain the empty vectors with selection markers BFP and puromycin for NDI1 and shRNA, respectively. NDI1-NDUFS3-A549 cells were resistant to the metformin-mediated reduction in cellular and mitochondrial oxygen consumption and cell proliferation (Figure 7—figure supplement 2B–G). Metformin also decreased HIF-1a protein levels in control but not in NDI1-NDUFS3-A549 cells (Figure 7—figure supplement 3A). Control-A549 cells subcutaneously injected into the left flank of nude mice rapidly grew in vivo, while tumors from mice fed metformin through drinking water ad libitum starting 2 weeks prior to tumor induction exhibited a marked reduction in growth over 45 days (Figure 7—figure supplement 3B). By contrast, NDI1-NDUFS3-A549 xenografts were completely resistant to metformin therapy, suggesting that metformin carries out its tumor inhibitory effects in a cell autonomous manner through inhibition of mitochondrial complex I in these cells (Figure 7—figure supplement 3C). NDI1-NDUFS3-A549 xenografts grew slower than control xenografts in untreated mice. An alternative explanation for the resistance of NDI1-NDUFS3-A549 cells to metformin could be that effects of metformin are blunted in slower-growing cells. However, based on our results from HCT116 cells in vivo and extensive analysis of A549 cells in vitro, we find that it is likely that NDI1 expressing A549 cells are also resistant to metformin in vivo due to rescue of complex I activity by the NDI1 protein.

The mechanisms by which metformin inhibits cancer growth is not fully understood. Metformin has been previously shown to inhibit mitochondrial complex I, yet it is not known whether metformin exhibits its anti-tumor effects through inhibition of complex I. It is important to note that many commonly utilized drugs have been shown to inhibit mitochondrial function in vitro; however, for many of these drugs it is not clear whether the in vivo mechanism of action is through modulating mitochondrial metabolism (Gohil et al., 2010). In the present study, we genetically demonstrate that metformin acts as a complex I inhibitor in vitro and in vivo to inhibit tumorigenesis. Our current findings bolster our previous findings highlighting the importance of mitochondrial metabolism as essential for cancer proliferation (Weinberg et al., 2010; Sullivan et al., 2013). Our results are consistent with a recent report demonstrating another biguanide phenformin exerts its anti-tumor effects by inhibiting complex I (Birsoy et al., 2014). However, unlike metformin, phenformin has been discontinued because of frequent occurrence of lactic acidosis (Bailey and Turner, 1996). While our study highlights the importance of mitochondrial complex I inhibition within cancer cells as a major mechanism by which metformin reduces tumor burden, it does not necessarily preclude any additional organismal effects of metformin that might reduce tumor progression in certain cancers. Specifically, cancer cells that express insulin receptors might be affected by metformin's inhibition on hepatic gluconeogenesis to reduce circulating insulin levels (Goodwin et al., 2012).

The levels of metformin within cells is regulated by a balance between uptake mechanisms, which are dependent on expression of organic cation transporters (OCT 1, 2, and 3), and expulsion mechanisms, which are dependent on expression on multidrug and toxin extrusion proteins (MATE1 and MATE 2) (Emami Riedmaier et al., 2013). Cancer cells are likely to have a wide range in expression of uptake and extrusion proteins allowing for metformin accumulation. The uptake of metformin in mitochondria is also dependent on mitochondrial membrane potential. Metformin exists as a cation at physiological pH and thus its accumulation within mitochondria is predicted to increase as a function of the mitochondrial membrane potential. The inhibition of electron transport at complex I by metformin should reduce the mitochondrial membrane potential as proton pumping is linked to electron transport. However, we observed that metformin-treated cells maintain their mitochondrial membrane potential in part due to the reversal of the ATP synthase thus allowing accumulation of metformin in the mitochondrial matrix (Figure 3). It is well known that cells with compromised ETC can sustain mitochondrial membrane potential through this mechanism (Appleby et al., 1999). Collectively, our results indicate that will inhibit mitochondrial complex I and decrease tumorigenesis in cancer cells that have transporters across the plasma membrane, as well as a robust inner mitochondrial membrane potential to allow metformin to reach the mitochondrial matrix. It will be of interest to observe whether metformin's efficacy as an anti-cancer agent is dependent on the tumor expression of OCTs.

Our study also reveals that metformin's inhibition on mitochondrial complex I is distinct from the classical complex I inhibitor rotenone. Metformin requires a robust mitochondrial membrane potential to accumulate in the mitochondrial matrix and reversibly inhibits complex I. By contrast, rotenone is an irreversible inhibitor of complex I that does not require mitochondrial membrane potential. Rotenone accumulation is not dependent on specific transporters expressed in the plasma membrane and readily accumulates in all cells and thus is highly toxic. Furthermore, metformin does not promote generation of ROS from complex I, while rotenone stimulates ROS production from complex I (Figure 5A). The flavin site within Complex I produces ROS and previous studies indicate that rotenone inhibits complex I downstream of the flavin site, stimulating ROS generation (Pryde and Hirst, 2011; St-Pierre et al., 2002). It is likely that metformin acts upstream of this site, inhibiting complex I activity while also inhibiting ROS generation. Mitochondrial complex III also produces ROS (Muller et al., 2004). Therefore, metformin by inhibiting the complex I would limit electron flow to complex III thereby reducing ROS generation from complex III. A consequence of mitochondrial complex III ROS generation is the hypoxic activation of HIF-1 (Bell et al., 2007). Indeed, we found that metformin reduced hypoxic stabilization of HIF-1α protein and HIF-dependent target genes. This result is consistent with our previous study in which we identified multiple mitochondrial inhibitors in a large-scale chemical screen for inhibitors of hypoxic activation of HIF (Lin et al., 2008). Beyond cancer, metformin might be an effective treatment for diseases associated with hyperactivation of HIF such as pulmonary hypertension (Shimoda and Semenza, 2011). The concentration of metformin we administered to mice in this report is predicted to achieve plasma concentrations of metformin similar to those in humans receiving metformin therapy. In our study, metformin was administered to drinking water at a concentration of 1.25 mg/ml, approximately 250 mg/kg. Conversion of doses from mice to human is achieved using a formula based on body surface area normalization (Reagan-Shaw et al., 2008). The human equivalent dose of 250 mg/kg metformin in mice is 20.27 mg/kg or 1418 mg for a 70 kg adult. Patients typically receive 500–2500 mg metformin per day (Scarpello and Howlett, 2008). In these patients receiving metformin therapy, plasma levels range between 0.5–2 µg/ml; approximately 4 µM–15 µM (Graham et al., 2011). Previous studies have shown that mice receiving drinking water containing 1–5 mg/ml of metformin have a plasma steady-state metformin concentration of 0.45–1.7 µg/ml (Memmott et al., 2010), indicating that our metformin dosage falls within a clinically relevant range.

Metformin concentrations (250–1000 µM) used for in vitro experiments in this report are in the range utilized by investigators in the diabetes field to examine the effects of metformin on primary murine and human hepatocytes (Foretz et al., 2010; Stephenne et al., 2011; Miller et al., 2013). A perplexing observation revealed in our study and previous studies is that the metformin concentrations required to induce biological effects in vitro are a magnitude of order higher than plasma level concentrations of metformin in vivo. A likely cause for this difference is the accumulation of metformin in liver or tumors to local concentrations that are much higher than in the circulating plasma. Indeed, the gut has been shown to accumulate metformin in the mM range (Bailey et al., 2008; Proctor et al., 2008).

In summary, our results indicate that metformin reversibly inhibits mitochondrial complex I within cancer cells to reduce tumorigenesis. Metformin inhibits tumorigenesis through multiple mechanisms including the induction of cancer cell death in conditions, when glucose is limited and through inhibition of mitochondrial ROS-dependent signaling pathways that promote tumorigenesis (i.e., HIF). These results indicate that metformin would be most effective in low glucose and oxygen conditions. It will be of interest to determine whether metformin treatment might provide a useful adjunct to therapies that limit glucose uptake (e.g., PI3K inhibitors) or drive tumors to low glucose and oxygen levels (e.g., anti-angiogenic inhibitors).

1. 1. Algire C
   2. Moiseeva O
   3. Deschenes-Simard X
   4. Amrein L
   5. Petruccelli L
   6. Birman E
   7. Viollet B
   8. Ferbeyre G
   9. Pollak MN

   (2012) Metformin reduces endogenous reactive oxygen species and associated DNA damage

   Cancer Prevention Research 5:536–543.

   https://doi.org/10.1158/1940-6207.CAPR-11-0536

   • Google Scholar
2. 1. Appleby RD
   2. Porteous WK
   3. Hughes G
   4. James AM
   5. Shannon D
   6. Wei YH
   7. Murphy MP

   (1999) Quantitation and origin of the mitochondrial membrane potential in human cells lacking mitochondrial DNA

   European Journal of Biochemistry 262:108–116.

   https://doi.org/10.1046/j.1432-1327.1999.00350.x

   • Google Scholar
3. 1. Bailey CJ
   2. Turner RC

   (1996) Metformin

   The New England Journal of Medicine 334:574–579.

   https://doi.org/10.1056/NEJM199602293340906

   • Google Scholar
4. 1. Bailey CJ
   2. Wilcock C
   3. Scarpello JH

   (2008) Metformin and the intestine

   Diabetologia 51:1552–1553.

   https://doi.org/10.1007/s00125-008-1053-5

   • Google Scholar
5. 1. Batandier C
   2. Guigas B
   3. Detaille D
   4. El-Mir MY
   5. Fontaine E
   6. Rigoulet M
   7. Leverve XM

   (2006) The ROS production induced by a reverse-electron flux at respiratory-chain complex 1 is hampered by metformin

   Journal of Bioenergetics and Biomembranes 38:33–42.

   https://doi.org/10.1007/s10863-006-9003-8

   • Google Scholar
6. 1. Bell EL
   2. Klimova TA
   3. Eisenbart J
   4. Moraes CT
   5. Murphy MP
   6. Budinger GR
   7. Chandel NS

   (2007) The Qo site of the mitochondrial complex III is required for the transduction of hypoxic signaling via reactive oxygen species production

   The Journal of Cell Biology 177:1029–1036.

   https://doi.org/10.1083/jcb.200609074

   • Google Scholar
7. 1. Birsoy K
   2. Possemato R
   3. Lorbeer FK
   4. Bayraktar EC
   5. Thiru P
   6. Yucel B
   7. Wang T
   8. Chen WW
   9. Clish CB
   10. Sabatini DM

   (2014) Metabolic determinants of cancer cell sensitivity to glucose limitation and biguanides

   Nature 508:108–112.

   https://doi.org/10.1038/nature13110

   • Google Scholar
8. 1. Birsoy K
   2. Sabatini DM
   3. Possemato R

   (2012) Untuning the tumor metabolic machine: targeting cancer metabolism: a bedside lesson

   Nature Medicine 18:1022–1023.

   https://doi.org/10.1038/nm.2870

   • Google Scholar
9. 1. Bowker SL
   2. Majumdar SR
   3. Veugelers P
   4. Johnson JA

   (2006) Increased cancer-related mortality for patients with type 2 diabetes who use sulfonylureas or insulin

   Diabetes Care 29:254–258.

   https://doi.org/10.2337/diacare.29.02.06.dc05-1558

   • Google Scholar
10. 1. Brand M

    (2010) The sites and topology of mitochondrial superoxide production

    Experimental Gerontology 45:466–472.

    https://doi.org/10.1016/j.exger.2010.01.003

    • Google Scholar
11. 1. Buzzai M
    2. Jones RG
    3. Amaravadi RK
    4. Lum JJ
    5. DeBerardinis RJ
    6. Zhao F
    7. Viollet B
    8. Thompson CB

    (2007) Systemic treatment with the antidiabetic drug metformin selectively impairs p53-deficient tumor cell growth

    Cancer Research 67:6745–6752.

    https://doi.org/10.1158/0008-5472.CAN-06-4447

    • Google Scholar
12. 1. Chandel N
    2. McClintock D
    3. Feliciano C
    4. Wood T
    5. Melendez J
    6. Rodriguez A
    7. Schumacker P

    (2000) Reactive oxygen species generated at mitochondrial complex III stabilize hypoxia-inducible factor-1alpha during hypoxia: a mechanism of O2 sensing

    The Journal of Biological Chemistry 275:25130–25138.

    https://doi.org/10.1074/jbc.M001914200

    • Google Scholar
13. 1. Dowling RJ
    2. Niraula S
    3. Stambolic V
    4. Goodwin PJ

    (2012) Metformin in cancer: translational challenges

    Journal of Molecular Endocrinology 48:R31–R43.

    https://doi.org/10.1530/JME-12-0007

    • Google Scholar
14. 1. El-Mir MY
    2. Nogueira V
    3. Fontaine E
    4. Averet N
    5. Rigoulet M
    6. Leverve X

    (2000) Dimethylbiguanide inhibits cell respiration via an indirect effect targeted on the respiratory chain complex I

    The Journal of Biological Chemistry 275:223–228.

    https://doi.org/10.1074/jbc.275.1.223

    • Google Scholar
15. 1. Emami Riedmaier A
    2. Fisel P
    3. Nies AT
    4. Schaeffeler E
    5. Schwab M

    (2013) Metformin and cancer: from the old medicine cabinet to pharmacological pitfalls and prospects

    Trends in Pharmacological Sciences 34:126–135.

    https://doi.org/10.1016/j.tips.2012.11.005

    • Google Scholar
16. 1. Evans JM
    2. Donnelly LA
    3. Emslie-Smith AM
    4. Alessi DR
    5. Morris AD

    (2005) Metformin and reduced risk of cancer in diabetic patients

    British Medical Journal 330:1304–1305.

    https://doi.org/10.1136/bmj.38415.708634.F7

    • Google Scholar
17. 1. Fendt SM
    2. Bell EL
    3. Keibler MA
    4. Davidson SM
    5. Wirth GJ
    6. Fiske B
    7. Mayers JR
    8. Schwab M
    9. Bellinger G
    10. Csibi A
    11. Patnaik A
    12. Blouin MJ
    13. Cantley LC
    14. Guarente L
    15. Blenis J
    16. Pollak MN
    17. Olumi AF
    18. Vander Heiden MG
    19. Stephanopoulos G

    (2013) Metformin decreases glucose oxidation and increases the dependency of prostate cancer cells on reductive glutamine metabolism

    Cancer Research 73:4429–4438.

    https://doi.org/10.1158/0008-5472.CAN-13-0080

    • Google Scholar
18. 1. Foretz M
    2. Hebrard S
    3. Leclerc J
    4. Zarrinpashneh E
    5. Soty M
    6. Mithieux G
    7. Sakamoto K
    8. Andreelli F
    9. Viollet B

    (2010) Metformin inhibits hepatic gluconeogenesis in mice independently of the LKB1/AMPK pathway via a decrease in hepatic energy state

    The Journal of Clinical Investigation 120:2355–2369.

    https://doi.org/10.1172/JCI40671

    • Google Scholar
19. 1. Gohil VM
    2. Sheth SA
    3. Nilsson R
    4. Wojtovich AP
    5. Lee JH
    6. Perocchi F
    7. Chen W
    8. Clish CB
    9. Ayata C
    10. Brookes PS
    11. Mootha VK

    (2010) Nutrient-sensitized screening for drugs that shift energy metabolism from mitochondrial respiration to glycolysis

    Nature Biotechnology 28:249–255.

    https://doi.org/10.1038/nbt.1606

    • Google Scholar
20. 1. Goodwin PJ
    2. Thompson AM
    3. Stambolic V

    (2012) Diabetes, metformin, and breast cancer: lilac time?

    Journal of Clinical Oncology 30:2812–2814.

    https://doi.org/10.1200/JCO.2012.42.3319

    • Google Scholar
21. 1. Graham GG
    2. Punt J
    3. Arora M
    4. Day RO
    5. Doogue MP
    6. Duong JK
    7. Furlong TJ
    8. Greenfield JR
    9. Greenup LC
    10. Kirkpatrick CM
    11. Ray JE
    12. Timmins P
    13. Williams KM

    (2011) Clinical pharmacokinetics of metformin

    Clinical Pharmacokinetics 50:81–98.

    https://doi.org/10.2165/11534750-000000000-00000

    • Google Scholar
22. 1. Hamanaka R
    2. Chandel N

    (2010) Mitochondrial reactive oxygen species regulate cellular signaling and dictate biological outcomes

    Trends in Biochemical Sciences 35:505–513.

    https://doi.org/10.1016/j.tibs.2010.04.002

    • Google Scholar
23. 1. Han D
    2. Antunes F
    3. Canali R
    4. Rettori D
    5. Cadenas E

    (2003) Voltage- dependent anion channels control the release of the superoxide anion from mitochondria to cytosol

    The Journal of Biological Chemistry 278:5557–5563.

    https://doi.org/10.1074/jbc.M210269200

    • Google Scholar
24. 1. Inzucchi SE
    2. Maggs DG
    3. Spollett GR
    4. Page SL
    5. Rife FS
    6. Walton V
    7. Shulman GI

    (1998) Efficacy and metabolic effects of metformin and troglitazone in type II diabetes mellitus

    The New England Journal of Medicine 338:867–872.

    https://doi.org/10.1056/NEJM199803263381303

    • Google Scholar
25. 1. Jamur MC
    2. Oliver C

    (2010) Permeabilization of cell membranes

    Methods in Molecular Biology 588:63–66.

    https://doi.org/10.1007/978-1-59745-324-0_9

    • Google Scholar
26. 1. Lin X
    2. David CA
    3. Donnelly JB
    4. Michaelides M
    5. Chandel NS
    6. Huang X
    7. Warrior U
    8. Weinberg F
    9. Tormos KV
    10. Fesik SW
    11. Shen Y

    (2008) A chemical genomics screen highlights the essential role of mitochondria in HIF-1 regulation

    Proceedings of the National Academy of Sciences of the United States of America 105:174–179.

    https://doi.org/10.1073/pnas.0706585104

    • Google Scholar
27. 1. Memmott RM
    2. Mercado JR
    3. Maier CR
    4. Kawabata S
    5. Fox SD
    6. Dennis PA

    (2010) Metformin prevents tobacco carcinogen–induced lung tumorigenesis

    Cancer Prevention Research 3:1066–1076.

    https://doi.org/10.1158/1940-6207.CAPR-10-0055

    • Google Scholar
28. 1. Miller RA
    2. Chu Q
    3. Xie J
    4. Foretz M
    5. Viollet B
    6. Birnbaum MJ

    (2013) Biguanides suppress hepatic glucagon signalling by decreasing production of cyclic AMP

    Nature 494:256–260.

    https://doi.org/10.1038/nature11808

    • Google Scholar
29. 1. Mullen AR
    2. Wheaton WW
    3. Jin ES
    4. Chen PH
    5. Sullivan LB
    6. Cheng T
    7. Yang Y
    8. Linehan WM
    9. Chandel NS
    10. DeBerardinis RJ

    (2012) Reductive carboxylation supports growth in tumour cells with defective mitochondria

    Nature 481:385–388.

    https://doi.org/10.1038/nature10642

    • Google Scholar
30. 1. Muller F
    2. Liu Y
    3. Van Remmen H

    (2004) Complex III releases superoxide to both sides of the inner mitochondrial membrane

    The Journal of Biological Chemistry 279:49064–49073.

    https://doi.org/10.1074/jbc.M407715200

    • Google Scholar
31. 1. Nathan DM
    2. Buse JB
    3. Davidson MB
    4. Ferrannini E
    5. Holman RR
    6. Sherwin R
    7. Zinman B
    8. American Diabetes Association
    9. European Association for Study of Diabetes

    (2009) Medical management of hyperglycemia in type 2 diabetes: a consensus algorithm for the initiation and adjustment of therapy: a consensus statement of the American Diabetes Association and the European Association for the Study of Diabetes

    Diabetes Care 32:193–203.

    https://doi.org/10.2337/dc08-9025

    • Google Scholar
32. 1. Ota S
    2. Horigome K
    3. Ishii T
    4. Nakai M
    5. Hayashi K
    6. Kawamura T
    7. Kishino A
    8. Taiji M
    9. Kimura T

    (2009) Metformin suppresses glucose-6-phosphatase expression by a complex I inhibition and AMPK activation-independent mechanism

    Biochemical and Biophysical Research Communications 388:311–316.

    https://doi.org/10.1016/j.bbrc.2009.07.164

    • Google Scholar
33. 1. Owen MR
    2. Doran E
    3. Halestrap AP

    (2000) Evidence that metformin exerts its anti-diabetic effects through inhibition of complex 1 of the mitochondrial respiratory chain

    The Biochemical Journal 348:607–614.

    https://doi.org/10.1042/0264-6021:3480607

    • Google Scholar
34. 1. Proctor WR
    2. Bourdet DL
    3. Thakker DR

    (2008) Mechanisms underlying saturable intestinal absorption of metformin

    Drug Metabolism and Disposition: the Biological Fate of Chemicals 36:1650–1658.

    https://doi.org/10.1124/dmd.107.020180

    • Google Scholar
35. 1. Pryde KR
    2. Hirst J

    (2011) Superoxide is produced by the reduced flavin in mitochondrial complex I: a single, unified mechanism that applies during both forward and reverse electron transfer

    The Journal of Biological Chemistry 286:18056–18065.

    https://doi.org/10.1074/jbc.M110.186841

    • Google Scholar
36. 1. Reagan-Shaw S
    2. Nihal M
    3. Ahmad N

    (2008) Dose translation from animal to human studies revisited

    FASEB Journal 22:659–661.

    https://doi.org/10.1096/fj.07-9574LSF

    • Google Scholar
37. 1. Rocha GZ
    2. Dias MM
    3. Ropelle ER
    4. Osorio-Costa F
    5. Rossato FA
    6. Vercesi AE
    7. Saad MJ
    8. Carvalheira JB

    (2011) Metformin amplifies chemotherapy-induced AMPK activation and antitumoral growth

    Clinical Cancer Research 17:3993–4005.

    https://doi.org/10.1158/1078-0432.CCR-10-2243

    • Google Scholar
38. 1. Scarpello JH
    2. Howlett HC

    (2008) Metformin therapy and clinical uses

    Diabetes & Vascular Disease Research 5:157–167.

    https://doi.org/10.3132/dvdr.2008.027

    • Google Scholar
39. 1. Semenza GL

    (2012) Hypoxia-inducible factors in physiology and medicine

    Cell 148:399–408.

    https://doi.org/10.1016/j.cell.2012.01.021

    • Google Scholar
40. 1. Sena LA
    2. Chandel NS

    (2012) Physiological roles of mitochondrial reactive oxygen species

    Molecular Cell 48:158–167.

    https://doi.org/10.1016/j.molcel.2012.09.025

    • Google Scholar
41. 1. Seo BB
    2. Kitajima-Ihara T
    3. Chan EK
    4. Scheffler IE
    5. Matsuno-Yagi A
    6. Yagi T

    (1998) Molecular remedy of complex I defects: rotenone-insensitive internal NADH- quinone oxidoreductase of Saccharomyces cerevisiae mitochondria restores the NADH oxidase activity of complex I-deficient mammalian cells

    Proceedings of the National Academy of Sciences of the United States of America 95:9167–9171.

    https://doi.org/10.1073/pnas.95.16.9167

    • Google Scholar
42. 1. Shackelford DB
    2. Abt E
    3. Gerken L
    4. Vasquez DS
    5. Seki A
    6. Leblanc M
    7. Wei L
    8. Fishbein MC
    9. Czernin J
    10. Mischel PS
    11. Shaw RJ

    (2013) LKB1 inactivation dictates therapeutic response of non-small cell lung cancer to the metabolism drug phenformin

    Cancer Cell 23:143–158.

    https://doi.org/10.1016/j.ccr.2012.12.008

    • Google Scholar
43. 1. Shimoda LA
    2. Semenza GL

    (2011) HIF and the lung: role of hypoxia-inducible factors in pulmonary development and disease

    American Journal of Respiratory and Critical Care Medicine 183:152–156.

    https://doi.org/10.1164/rccm.201009-1393PP

    • Google Scholar
44. 1. St-Pierre J
    2. Buckingham JA
    3. Roebuck SJ
    4. Brand MD

    (2002) Topology of superoxide production from different sites in the mitochondrial electron transport chain

    The Journal of Biological Chemistry 277:44784–44790.

    https://doi.org/10.1074/jbc.M207217200

    • Google Scholar
45. 1. Stephenne X
    2. Foretz M
    3. Taleux N
    4. van der Zon GC
    5. Sokal E
    6. Hue L
    7. Viollet B
    8. Guigas B

    (2011) Metformin activates AMP-activated protein kinase in primary human hepatocytes by decreasing cellular energy status

    Diabetologia 54:3101–3110.

    https://doi.org/10.1007/s00125-011-2311-5

    • Google Scholar
46. 1. Sullivan LB
    2. Martinez-Garcia E
    3. Nguyen H
    4. Mullen AR
    5. Dufour E
    6. Sudarshan S
    7. Licht JD
    8. Deberardinis RJ
    9. Chandel NS

    (2013) The proto-oncometabolite fumarate binds glutathione to amplify ROS-dependent signaling

    Molecular Cell 51:236–248.

    https://doi.org/10.1016/j.molcel.2013.05.003

    • Google Scholar
47. 1. Tomimoto A
    2. Endo H
    3. Sugiyama M
    4. Fujisawa T
    5. Hosono K
    6. Takahashi H
    7. Nakajima N
    8. Nagashima Y
    9. Wada K
    10. Nakagama H
    11. Nakajima A

    (2008) Metformin suppresses intestinal polyp growth in ApcMin/+ mice

    Cancer Science 99:2136–2141.

    https://doi.org/10.1111/j.1349-7006.2008.00933.x

    • Google Scholar
48. 1. Viollet B
    2. Guigas B
    3. Sanz Garcia N
    4. Leclerc J
    5. Foretz M
    6. Andreelli F

    (2012) Cellular and molecular mechanisms of metformin: an overview

    Clinical Science 122:253–270.

    https://doi.org/10.1042/CS20110386

    • Google Scholar
49. 1. Vogel RO
    2. Dieteren CE
    3. van den Heuvel LP
    4. Willems PH
    5. Smeitink JA
    6. Koopman WJ
    7. Nijtmans LG

    (2007) Identification of mitochondrial complex I assembly intermediates by tracing tagged NDUFS3 demonstrates the entry point of mitochondrial subunits

    The Journal of Biological Chemistry 282:7582–7590.

    https://doi.org/10.1074/jbc.M609410200

    • Google Scholar
50. 1. Weinberg F
    2. Hamanaka R
    3. Wheaton W
    4. Weinberg S
    5. Joseph J
    6. Lopez M
    7. Kalyanaraman B
    8. Mutlu G
    9. Budinger G
    10. Chandel N

    (2010) Mitochondrial metabolism and ROS generation are essential for Kras-mediated tumorigenicity

    Proceedings of the National Academy of Sciences of the United States of America 107:8788–8793.

    https://doi.org/10.1073/pnas.1003428107

    • Google Scholar

Author details

1. William W Wheaton

   Department of Medicine, The Feinberg School of Medicine, Northwestern University, Chicago, United States

   Contribution

   WWW, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Contributed equally with

   Samuel E Weinberg

   Competing interests

   The authors declare that no competing interests exist.

2. Samuel E Weinberg

   Department of Medicine, The Feinberg School of Medicine, Northwestern University, Chicago, United States

   Contribution

   SEW, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Contributed equally with

   William W Wheaton

   Competing interests

   The authors declare that no competing interests exist.

3. Robert B Hamanaka

   Department of Medicine, The Feinberg School of Medicine, Northwestern University, Chicago, United States

   Contribution

   RBH, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

4. Saul Soberanes

   Department of Medicine, The Feinberg School of Medicine, Northwestern University, Chicago, United States

   Contribution

   SS, Conception and design, Acquisition of data, Analysis and interpretation of data

   Competing interests

   The authors declare that no competing interests exist.

5. Lucas B Sullivan

   Department of Medicine, The Feinberg School of Medicine, Northwestern University, Chicago, United States

   Contribution

   LBS, Conception and design, Acquisition of data, Analysis and interpretation of data

   Competing interests

   The authors declare that no competing interests exist.

6. Elena Anso

   Department of Medicine, The Feinberg School of Medicine, Northwestern University, Chicago, United States

   Contribution

   EA, Conception and design, Acquisition of data, Analysis and interpretation of data

   Competing interests

   The authors declare that no competing interests exist.

7. Andrea Glasauer

   Department of Medicine, The Feinberg School of Medicine, Northwestern University, Chicago, United States

   Contribution

   AG, Conception and design, Acquisition of data, Analysis and interpretation of data

   Competing interests

   The authors declare that no competing interests exist.

8. Eric Dufour

   Institute of Biomedical Technology, University of Tampere, Tampere, Finland

   Contribution

   ED, Conception and design, Contributed unpublished essential data or reagents

   Competing interests

   The authors declare that no competing interests exist.

9. Gokhan M Mutlu

   Department of Medicine, The Feinberg School of Medicine, Northwestern University, Chicago, United States

   Contribution

   GMM, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

10. GR Scott Budigner

    Department of Medicine, The Feinberg School of Medicine, Northwestern University, Chicago, United States

    Contribution

    GRSB, Conception and design, Analysis and interpretation of data, Drafting or revising the article

    Competing interests

    The authors declare that no competing interests exist.

11. Navdeep S Chandel

    Department of Medicine, The Feinberg School of Medicine, Northwestern University, Chicago, United States

    Contribution

    NSC, Conception and design, Analysis and interpretation of data, Drafting or revising the article

    For correspondence

    nav@northwestern.edu

    Competing interests

    The authors declare that no competing interests exist.

• 12,204

  views

• 2,067

  downloads

• 851

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.