How to starve cancer cells when nutrients are abundant

ABSTRACT

Metabolic flexibility represents a potential point of attack for novel cancer treatments. We recently described the signaling mechanism inducing a metabolic shift in response to metformin and phenformin in leukemia and lymphoma cells. Enhanced glucose utilization was critically dependent on mitochondrial stress signaling/hypoxia-inducible factor-1α representing a therapeutic vulnerability.

Most people would agree that flexibility is an overall advantageous trait. From economics to evolutionary biology, flexibility and adaptability are viewed as critical elements of fitness and survival. In cancer, this essential trait, however, represents a major obstacle on the way of finding a successful cure. Molecular targeted therapies such as small-molecule BRAF inhibitors for malignant melanoma show initial responses in a significant number of patients before cancer cells ultimately develop resistance through rewiring their signaling networks enabling proliferation and survival.

For the development of drugs targeting metabolic pathways, we should learn from this experience and define targets representing metabolic “bottlenecks” in cancer cells. Cellular metabolism is a highly complex and dynamic network responding to numerous cell intrinsic and extrinsic regulating factors allowing adaptation to various nutrient sources and environments. Cancer cells are able to utilize various forms of fuels such as glucose, amino acids or lactate to support proliferation with metabolic signatures and preferences depending on genetic background and nutrient microenvironment (Figure 1a).¹ Defining metabolic treatment targets, therefore, critically depends on identification of limiting metabolic requirements. Another, albeit not mutually exclusive, approach would be to target metabolic flexibility to prevent compensatory adaptation to metabolic drugs. Limiting availability of nutrients in the tumor microenvironment that have become critical during metabolic adaptation through pharmacologic or dietary means such as various forms of fasting is an area of ongoing research.² Another approach which we explored in our work is to identify rational drug combinations that block signaling pathways leading to adaptive metabolic reprogramming in response to metabolic drugs.³

Figure 1.

Compensatory increase in glucose utilization upon exposure to biguanides can be blocked by inhibition of the mROS/HIF-1α axis. (a) Cancer cells are able to metabolize various fuels driven by numerous extracellular and intracellular factors. (b) Inhibition of mitochondrial complex I through biguanides metformin and phenformin leads to metabolic reprogramming with increased glucose utilization and aerobic glycolysis mediated by mitochondrial stress signaling through mROS/HIF-1α. (c) Suppression of mitochondrial stress signaling prevents metabolic adaptation and compensatory glucose utilization causing synthetic lethality under nutrient-rich conditions (“starvation”). ETC, electron transport chain; HIF-1α, hypoxia-inducible factor-1α; mROS, mitochondrial reactive oxygen species; TCA cycle, tricarboxylic acid cycle.

Today’s intense effort to identify unique metabolic pathways as potential therapeutic targets for cancer goes back to work performed at the Kaiser Wilhelm Institute (KWI) for Biology in Berlin, Germany, about 100 years ago. There, Otto Warburg made the seminal discovery that cancer cells produce large amounts of lactate under aerobic conditions which is now known as the “Warburg effect”.⁴ Since then our understanding of cancer cell metabolism has grown exponentially and, against Warburg’s conviction, mitochondrial function has been shown to be critically important in tumorigenesis and cancer progression explaining the current interest in mitochondria as a cancer drug target.⁵ The most recent evidence comes from studies showing the importance of mitochondrial electron transport chain and tricarboxylic acid (TCA) cycle in cancer cell proliferation through generation of aspartate, a critical amino acid for macromolecule synthesis and a limiting requirement for cellular proliferation.⁶

Another factor contributing to the current interest in mitochondria as cancer drug targets stems from observation in patients with type 2 diabetes (T2D). Initially discovered for its blood glucose-lowering effect, the biguanide metformin is now the most commonly prescribed drug for patients with type 2 diabetes. Epidemiological studies have suggested that diabetic patients on metformin show a reduced risk of cancer igniting the current efforts to study its potential antineoplastic effect and a focus on the therapeutic potential of targeting mitochondrial metabolism in cancer.⁷ Through inhibition of complex I of the mitochondrial electron transport chain, metformin and the related drug phenformin show anticancer effects in cell culture and animal models.⁸ Importantly, the antineoplastic effect of metformin and phenformin shows marked variation between cancer models, different genetic backgrounds, and nutrient availability all of which provide the basis for metabolic plasticity during biguanide-induced mitochondrial stress.⁹

In our recent work, we characterized metabolic reprogramming and identified cancer-specific metabolic vulnerabilities in response to pharmacological inhibition of mitochondrial complex I in leukemia and lymphoma cells. Resistance against mitochondrial dysfunction induced by biguanides depends on reprogramming of glucose metabolism to activate aerobic glycolysis. As part of this metabolic rewiring, cells narrowed their fuel preferences by increasing glucose uptake and utilization and becoming highly dependent on it (Figure 1b). We could show that metabolic adaptation to complex I dysfunction is mediated by mitochondrial reactive oxygen species (mROS) serving as a stress signal that activates hypoxia-inducible factor-1a (HIF-1α). Inhibition of the mROS/HIF-1α axis through either antioxidants or suppression of HIF-1α selectively disrupts metabolic adaptation and survival. Combined inhibition of mitochondrial complex I and mROS/HIF-1α signaling disabled metabolic flexibility and prevented adaptive glucose utilization independent of its availability (“starvation”; Figure 1c)

The translational value of this observation lies in the fact that the mROS/HIF-1α-dependent metabolic reprogramming was confirmed in acute lymphoblastic leukemia (ALL) and chronic lymphocytic leukemia (CLL) suggesting novel therapeutic strategies for lymphoid leukemias. Moreover, the observed rewiring through mROS/HIF-1a seemed to be selective for malignant lymphocytes. Normal lymphocytes isolated from healthy donors did not show activation of or dependence on the HIF-1α pathway during biguanide exposure.

This work illustrates the challenge in targeting metabolism in cancer cells. Inhibition of mitochondrial complex I with biguanides induced a glucose-dependent state similar to that observed previously.^8,10 As long as nutrients and in particular glucose were abundant, malignant lymphocytes could survive and grow in spite of diminished oxidative phosphorylation by rewiring their metabolic profile to increase glucose uptake and conversion into lactate (Figure 1b). In this case, rational drug combinations represent a way toward targeting the highly dynamic metabolic network of cancer cells. Our results identify HIF-1α signaling as a critical factor in the resistance against metformin- and phenformin-induced mitochondrial dysfunction allowing selective targeting of metabolic pathways in lymphoid leukemia helping to identify novel targets for metabolic interventions.

Funding Statement

This work was supported by University of Wisconsin Skin Disease Research Center Pilot Grant, University of Wisconsin Carbone Cancer Center Support Grant (#P30 CA014520),American Cancer Society Institutional Research Grant (#86-004-26), and a Dermatology Foundation Medical Dermatology Career Development Award.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.