Date Published: November 12, 2011
Publisher: Impact Journals LLC
Author(s): Javier A. Menendez, Sílvia Cufí, Cristina Oliveras-Ferraros, Begoña Martin-Castillo, Jorge Joven, Luciano Vellon, Alejandro Vazquez-Martin.
By activating the ataxia telangiectasia mutated (ATM)-mediated DNA Damage Response (DDR), the AMPK agonist metformin might sensitize cells against further damage, thus mimicking the precancerous stimulus that induces an intrinsic barrier against carcinogenesis. Herein, we present the new hypothesis that metformin might function as a tissue sweeper of pre-malignant cells before they gain stem cell/tumor initiating properties. Because enhanced glycolysis (the Warburg effect) plays a causal role in the gain of stem-like properties of tumor-initiating cells by protecting them from the pro-senescent effects of mitochondrial respiration-induced oxidative stress, metformin’s ability to disrupt the glycolytic metabotype may generate a cellular phenotype that is metabolically protected against immortalization. The bioenergetic crisis imposed by metformin, which may involve enhanced mitochondrial biogenesis and oxidative stress, can lower the threshold for cellular senescence by pre-activating an ATM-dependent pseudo-DDR. This allows an accelerated onset of cellular senescence in response to additional oncogenic stresses. By pushing cancer cells to use oxidative phosphorylation instead of glycolysis, metformin can rescue cell surface major histocompatibility complex class I (MHC-I) expression that is downregulated by oncogenic transformation, a crucial adaptation of tumor cells to avoid the adaptive immune response by cytotoxic T-lymphocytes (CTLs). Aside from restoration of tumor immunosurveillance at the cell-autonomous level, metformin can activate a senescence-associated secretory phenotype (SASP) to reinforce senescence growth arrest, which might trigger an immune-mediated clearance of the senescent cells in a non-cell-autonomous manner. By diminishing the probability of escape from the senescence anti-tumor barrier, the net effect of metformin should be a significant decrease in the accumulation of dysfunctional, pre-malignant cells in tissues, including those with the ability to initiate tumors. As life-long or late-life removal of senescent cells has been shown to prevent or delay the onset or progression of age-related disorders, the tissue sweeper function of metformin may inhibit the malignant/metastatic progression of pre-malignant/senescent tumor cells and increase the human lifespan.
Energetic stress due to glucose restriction increases the AMP/ATP ratio. Treatments with drugs that increase the AMP/ATP ratio, including the AMP analog 5-aminoimidazole-4-carboxamide-1-β-ribofuranoside (AICAR) or the anti-diabetic biguanide metformin, activate AMPKα through phosphorylation of Thr-172 and also increase the levels of the AMPKα protein. Although several proteins can phosphorylate AMPKα (e.g., the master upstream Ser/Thr kinase 11 (STK11)/Liver Kinase B1 [LKB1]), it should be noted that activating phosphorylation of AMPKα in response to energetic stress takes place in an ATM-dependent and STK11/LKB1-independent manner . Accordingly, the selective ATM inhibitor KU-55933 markedly reduces the AMPKα-activating effects of metformin in rat hepatoma cells, functionally supporting the first genome-wide association study that unexpectedly found the ATM gene as the causal modulator of glycemic responsiveness to metformin among type 2 diabetic patients . Indeed, treatment with the ATM inhibitor KU-55933 is sufficient to prevent metformin-induced phosphorylation of AMPKα and of the AMPKα downstream target Acetyl-CoA Carboxylase (ACC), concluding that ATM works upstream of AMPKα and that ATM is required for a full response to metformin . Although these results support and extend previous reports of ATM involvement in the activation of AMPKα by stimuli other than metformin [7, 9, 10], metformin’s ability to function as a general activator of the ATM-dependent DDR pathway remains to be explored to prove a causal link between the metformin-induced activation of ATM and the diminished risk of developing cancer in individuals taking this drug .
As metformin is thought to activate AMPK by inhibiting oxidative phosphorylation [18, 19] and because phosphorylation of CREB at Ser-133 can be observed in cultured cells that have been incubated with oxidative phosphorylation inhibitors , it could be argued that metformin-induced CREB activation might merely reflect metformin’s ability to impair mitochondrial activity in tumor cells. However, CREB phosphorylation is pivotal in mediating peroxisome proliferator-activated receptor gamma coactivator-1alpha (PGC-1α)-stimulated mitochondrial biogenesis [21-24]. Therefore, metformin-stimulated phosphorylation of CREB at Ser-133, which activates the promoter of PGC-1α and increases PGC-α mRNA and protein expression [25, 26], can also be viewed as part of the mechanism through which metformin may control mitochondrial biogenesis in tumor cells. Tumor cells are dependent on glycolysis to support their metabolic requirements; even under aerobic conditions, tumor cells continue to rely on glycolysis rather than oxidative phosphorylation (Warburg effect), resulting in high glucose requirements to generate energy and biosynthetic precursors because of the increased availability of glycolytic intermediates [27-30]. As such, metformin-induced reactivation of oxidative phosphorylation biogenesis may contribute to the growth arrest of cancer cells. A recently developed high-throughput respirometric assay for mitochondrial biogenesis used the Seahorse Bioscience analyzer to measure mitochondrial function in real time. In adapted primary cultures of non-glycolytic renal proximal tubular cells, metformin augmented mitochondrial biogenesis . Recent experiments from our own laboratory have established that culturing human cancer cells in the presence of metformin significantly enhances the expression of cytochrome c oxidase I (COX-1) and mitochondrial succinate dehydrogenase (SDH-A), which are encoded by mitochondrial and nuclear genomes, respectively (Oliveras-Ferraros C, Cufí S, Vazquez-Martin A, Menendez OJ, Martin-Castillo B, Joven J, Menendez JA. Metformin rescues cell surface major histocompatibility complex class I deficiency caused by oncogenic transformation. Submitted for publication). Using cancer cell lines, non-cancer cells, embryonic cells and Rho(0) cells (i.e., cells depleted of mitochondrial DNA), Jose et al.  recently confirmed that the AMPK agonist AICAR exhibits a strong and cancer-specific growth effect that depends on the bioenergetic signature of the cells and involves upregulation of oxidative phosphorylation. In fact, the sensitivity to pharmacological activation of AMPK is higher when cells display a high proliferation rate accompanied by a low steady-state content of ATP. Although it remains to be established if AMPKα-related induction of mitochondrial biogenesis to increase oxidative phosphorylation is instrumental and possibly required for the anti-cancer/anti-aging effects of metformin [33, 34], it is becoming clear that some health-promoting capabilities of metformin may rely on its ability to function as a bona fide glucose-starvation mimetic. Accordingly, recent experiments from our own laboratory have confirmed that, when added to glucose-free medium, where growth is highly oxidative phosphorylation-dependent, metformin drastically increases apoptotic cell death in glucose-addicted cancer cell cultures. Because transformed human cell types appear to be more sensitive to glucose deprivation-induced cytotoxicity and metabolic oxidative stress than non-transformed human cell types, we suggest that a rational use of metformin in combination with fasting could significantly potentiate the effects of chemotherapy in cancer while protecting normal cells, thus further increasing the therapeutic window [35-38] (Oliveras-Ferraros C, Cufí S, Vazquez-Martin A, Menendez OJ, Joven J, Martin-Castillo B, Menendez JA. Glucose deprivation enhances metformin-induced apoptosis in a breast cancer cell type-dependent manner: Implications for cyclotherapy. Manuscript in preparation).
Understanding the metabolic changes associated with somatic cell reprogramming might shed light on the “metabolic transformation” that is required to support not only the increased biosynthetic needs of the tumor cell but also to enable the acquisition of stemness properties in cancer stem cells (CSCs). Many of the changes in cell metabolism that have been identified to be important in regulating somatic cell reprogramming and induced pluripotency also play roles in oncogenesis. On the other hand, reprogramming to a more dedifferentiated state occurs during tumor progression (i.e., the activation of an embryonic stem cell-like transcriptional program in differentiated adult cells may induce pathologic self-renewal/stemness characteristics of CSCs ) and might be favored by alterations in crucial tumor suppressors. Indeed, the stress mechanisms triggered by expression of the Yamanaka stemness factors, which ultimately lead to reprogramming, elicit the tumor suppressor pathways that naturally protect cells against the uncontrolled growth that occurs during tumorigenesis. Unsurprisingly, the cellular response to expression of the reprogramming factors or stem-cell-specific genes molecularly mimics the senescence response observed during OIS, thus emphasizing the parallels between RIS and OIS. In this scenario, if metformin treatment appears to reinforce RIS in somatic reprogramming experiments, we can then infer that metformin treatment should improve cells’ ability to establish a more efficient senescence response in pre-malignant and malignant tissues. Many tumor cells appear to have developed mechanisms to reduce AMPK activation and therefore escape from its growth-arrest and tumor-suppressor effects. In fact, more aggressive tumors exhibit reduced signaling via the AMPK pathway, and an inverse relationship exists between the AMPK activation status with histological grade and metastasis [92-94]. As such, metformin-induced energy crisis and, therefore, AMPK (re-)activation, may functionally disrupt the deleterious connection between pluripotency and oncogenic transformation. In fact, Struhl’s team discovered that metformin treatment can selectively kill the chemotherapy-resistant subpopulation of CSCs in genetically distinct types of breast cancer cell lines [95, 96]. Our own group has confirmed that treatment with metformin can suppress the self-renewal and proliferation of cancer stem/progenitor cells in HER2 gene-amplified breast carcinomas cells refractory to HER2-targeted drugs [97, 98]. The central mechanisms through which metformin exposure blocks the ontogenesis of the CSC molecular signature have begun to be elucidated in cultured cancer cells, including alterations of epithelial-to-mesenchymal transition drivers/effectors, tumor-suppressor miRNAs and oncomiRs [99-101]. Metformin’s ability to oppose reprogramming of cell energy metabolism from oxidative mitochondria toward an alternative ATP-generating glycolytic metabotype may be sufficient to inhibit the network required for the establishment and maintenance of stem cell pluripotency and self-renewal imposed by certain oncogenic stimuli in the right cellular context .
Metformin’s ability to enhance senescence in established premalignant disease or in fully malignant disease is a largely unexplored mechanism that may explain why reductions in cancer mortality related to metformin use are similar in magnitude to reductions in cancer incidence. This suggests that the anti-cancer effects of metformin largely depend on (or are restricted to) its preventive effects . The most widely accepted interpretation for the biological function of cellular senescence is that it serves as a mechanism for restricting cancer progression. Based on this, escaping from cellular senescence and becoming immortal constitute a required additional step in the progression of oncogenesis [51, 52]. Recent studies have suggested that the accumulation of ROS and oxidative damage are commonly involved in culture stress- or oncogene-induced cellular senescence. Increasing accumulation of ROS is observed during replicative senescence, and the replicative potential of MEFs and HDFs is significantly higher under low oxygen conditions. As such, the ability of immortalized cells, including embryonic stem cells (ESCs), iPSCs and CSCs, to buffer oxidative stress may be pivotal for explaining their immortality [53-56; 87-90]. Enhanced glycolysis actively protects cells from senescence induced by oxidative stress [53-56], a metabolic protection that appears to causally contribute to the maintenance of the self-renewal capacity of stem cells [87-90]. In fact, the enhanced glycolysis of the Warburg effect is a crucial metabolic feature that helps cancer cells bypass senescence, and this may provide indirect evidence that metformin’s primary target is the immortalizing step during tumorigenesis. In other words, if enhanced glycolysis is necessary and sufficient to enable indefinite proliferation (i.e., immortalization) very early during multi-step carcinogenesis in vivo, then metformin’s ability to inhibit glucose flux while simultaneously stimulating lactate/pyruvate flux and mitochondrial biogenesis must cause ATP depletion accompanied by a drastic increase in cellular AMP, which is expected to induce premature senescence . Many tumor cells retain the ability to senesce in response to DNA-damaging drugs in culture and in vivo. Because of this, metformin-accelerated replicative senescence due to a stronger DDR-dependent cell cycle arrest may underlie metformin’s ability to increase the rate of pathological complete response (pCR) in neoadjuvant chemotherapy in diabetic patients with breast cancer  and to promote tumor regression and prevent relapse when combined with suboptimal doses of chemotherapy in animal models .
In addition to SA-β-gal activity, senescence has been previously linked to induction of ROS production, which is believed to be necessary for maintenance of the senescence phenotype [139, 140]. ROS also play a pivotal role in promoting SA-β-gal activity following radiation-induced DNA damage. Accordingly, inhibition of ROS using N-acetyl cysteine (NAC) following radiation treatment drastically decreases senescence in tumor cells with normally high levels of radiation-induced SA-β-gal. In accordance with our current observations and taking advantage of earlier studies suggesting that metformin treatment induces ROS in certain cellular backgrounds , Skinner et al  have recently confirmed that enhanced rather than reduced ROS and SA-β-gal activity occurred when metformin was concurrently added to radiation in p53-deficient tumor cells. Skinner’s findings that metformin treatment can overcome locoregional treatment failure in head and neck carcinomas by reinforcing the radiation-induced cellular senescence, which clinically translates into a significantly improved survival in patients taking metformin, strongly support our hypothesis that metformin’s ability to reinforce the establishment of accelerated senescence may function as an effective barrier to tumor growth and disease recurrence.