In the opening plenary session of the 2007 meeting of the American Association of Cancer Research (AACR07), Craig B. Thompson, of the University of Pennsylvania, presented the provocative proposal that most oncogenes and tumor suppressors evolved to regulate cellular metabolism. In particular, oncogenes and tumor suppressors modulate the cellular response to growth factor-derived signals that direct nutrient uptake; and they enable cells to adapt intermediate metabolism to fuel cellular responses to stress in bioenergetic or biosynthetic pathways.
- The rising cancer rate associated with obesity will be shown to result from carbohydrate-driven increases in IGF-1 production, and an increase in this fuel signal that drives spontaneous mutation in all epithelial tumors. Mitochondrial-directed reactive oxygen species (ROS), resulting from excessive oncogene-directed nutrient uptake, will be determined to be the major cause of the highest mutation rate in cancer and explains the vast majority of human cancer mutations.
- If the hypothesis above is correct then effective cancer preventive agents may already be available. The first widely used cancer prevention agents will be either mitochondrial uncoupling agents such as TNP (2,4,6-trinitrophenol) and AMP activators such as metformin currently used in the treatment of Type 2 diabetes. Four international studies suggest that Type 2 diabetics treated with metformin have a reduced cancer risk than non-treated diabetics.
- A "Warbug-like" uptake and metabolism of amino acids by tumors will be established as the cause of cancer cachexia.
- Companies are investigating drugs that impair glycolytic metabolism or fatty acid synthesis. New drugs will be introduced, joining existing nucleoside analogs and protein synthesis inhibitors as cancer therapeutics. Such agents are likely to be safe because only growing cells are subjected to glucose withdrawal-induced death. Most other cells are vegetative and can survive that stress. New fatty acid and protein synthesis inhibitors are under study.
- Antagonists of anti-apoptotic Bcl-2 proteins will prove most effective in PET scan-negative tumors.
- Inhibitors of autophagy will become major components of most chemotherapeutic regimens.
- Successful targeted therapies will act by suppressing the primary abnormality in cancer, i.e. cell autonomous nutrient uptake.
Theory, Mechanisms, and Supporting Data
Craig Thompson presented the provocative proposal that most oncogenes and tumor suppressors evolved to regulate cellular metabolism. In mammalian biology, nutrient uptake and metabolism is one of the major processes that the signal transduction tyrosine kinase signaling pathway has evolved to regulate. This statement may seem bold because, despite recent updates in charts of pathways in human cancer (Weinberg, 2006) and metabolic pathways (Nicholson, 2007), not a single protein overlaps both pathways. Why should we then consider this proposal?
The traditional view holds that receptor-initiated cell proliferation regulates metabolism indirectly. After signal transduction, new transcription and translation drives the process. A decrease in ATP drives increased rate of glucose metabolism, and an increase in ADP drives oxidative phosphorylation. However, mammalian cells require extracellular signal transduction to transcribe nutrient transporters for survival. Normal, non-transformed cells that lose this ability by dislodging or losing hormonal support start an apoptotic clock, losing necessary bioenergetics and mitochondrial function.
The loss of required bioenergetics in cells is sensed by proapoptic bcl-2 family members, which initiate apoptosis within 48-96 hours. During neoplastic transformation, cells gain anti-bcl-2 proteins that can keep them alive under this paradigm. Their survival, however, depends not on regaining the ability to take up nutrients to maintain nutrient and mitochondrial physiology, but on gaining the ability to further self-catabolize additional cell contents through a process known as autophagy. Transformed cells adapt a stress form of cell survival that allows them to live for 3-4 weeks, during which they can migrate to a new site where they receive signal transduction or acquire additional mutation events, providing new means of becoming cell autonomous. [For normal, non-transformed cells, the loss of the ability to receive signal transduction from growth factors that regulates nutrient transporter expression is an irrevocable mechanism of death during ensuing days.]
This observation leads to a hypothesis that optimal cell proliferation requires two independent transformation events for a cancer cell to become cell autonomous for cell proliferation, a growth signal and a fuel signal. The growth signal is well appreciated by cancer researchers. It is signal transduction that informs new transcription and translation, which is an energy dependent process. To be effective, however, the growth signal must be matched by other signal transduction to stimulate the cell to take up the nutrients to meet the requirements for growth, which include both ATP production (to meet energy requirements) and making available the building blocks for ensuing growth and proliferation.
Because it is unlikely that simultaneous mutations occur in a cancer cell, the question emerges, which one occurs first, the growth signal or the fuel signal? Consider either signal alone:
- In the case of perturbed growth signals, cells transformed only with an activated component of growth signaling die. This scenario is best characterized by the oncogene, myc. Myc is effectively apoptotic, acting by consumption of remaining ATP by inducing new transcription and translation. This mechanism is effective for protecting mammals from activating mutations in signal transduction pathways that render cell growth and proliferation cell autonomous.
- In contrast, dysregulation in fuel signaling, alone, might have the opposite effect. Enhanced fuel signaling alone would promote survival accompanied by reactive oxygen species (ROS) production and DNA mutation.
Dysregulations in fuel signaling, alone, are likely to be the first mutation events in the pathway to somatic genetic changes that lead to cancer. In this case, increased uptake of glucose would fuel increased ATP production, allowing the cell to increase transcription and translation rates and defeat the apoptotic pathway, allowing the cell to become cell autonomous for survival. The development of cell autonomy could provoke an interesting secondary consequence at the mitochondria. As the cell takes up more energy than it needs and metabolizes it, mitochondrial generated ROS is formed, causing Fenton reactions that damage DNA. Dysregulated fuel signaling, then, would have the potential not only to increase the survival of a cell that sustained such a lesion, but also to increase the mutation rate of that cell.
One problem of this theory is that very little evidence during the last 100 years supports the existence of such a pathway. However, during the last 10 years, support has emerged, including studies of glucose metabolism in malignancy and clinical reliance on a test based on Warburg's observations that tumor cells have an abnormally high glycolytic rate. In the clinic, PET scan imaging of tumors takes advantage of high rate of glucose metabolism because malignant cells outcompete normal cells for glucose uptake. The labeled glucose analog, 18-fluoro-2-deoxyglucose is taken up and phosphorylated by cells in a manner similar to glucose. Although this phenomenon has never been fully explained, PET scanning predicts poor prognosis and allows detection of metastases earlier than most other methods.
Human and cancer genome research provides a likely molecular explanation. The most common activating mutation in all spontaneous human tumors are activating mutations of the p110 catalytic subunit of PI3K, which occurs in 25% of epithelial malignancies, and is involved in an insulin-like signaling pathway in every cell of our body. The net consequences are regulation of glucose uptake and metabolism. The pathway initiates in every cell type studied to date, with RTK initiating the activation of PI3K, to produce PIP3, which recruits to the membrane and activates two serial kinases, PDK-1 and AKT.
In both the metabolic and cancer research, Akt activation has been shown to be necessary and sufficient to regulate cellular glucose uptake (i.e., constitutively active Akt transgene is sufficient to stimulate glycolysis). This occurs by regulating glucose transporter function (surface expression), hexokinase activity (by phosphorylation) to capture glucose intracellularly, and PFK-1 activation to initiate catabolism of glucose into its elemental constituents to be used for energy consumption and biomolecular synthesis. If activated Akt is introduced into non-transformed cell lines in culture, cells spontaneously switch over to aerobic glycolysis, exactly as Warburg described it 80 years ago. In an AKT-dependent fashion, they take up more glucose from their environment and suppress oxidative phosphorylation. Excessive secretion of lactate by the epithelial and hematopoietic cells that form the bulk of cancer and do not have storage capacity for excess carbon, explains the mechanism of tumor imaging by PET scan.
Why is the drive to aerobic glycolysis adaptive for so many tumor types, particularly poor prognosis tumors? It is important to recognize that Akt contributes beyond glucose uptake to support tumor cell adaptation. One way is via a new pathway by which glucose can be converted into the fatty acids required to produce 2 daughter cells, through the production of cholesterol and elongating acyl chains. Akt phosphorylation of ATP citrate lyase (ACL) results in glucose-dependent lipid synthesis through downstream substrates that switch mitochondrial metabolism. This pathway leads phospholipid biosynthesis, which is required for production of the biomembranes that the new daughter cells need.
Another Akt contribution to tumor cell adaptation occurs when high glucose input and Akt activation, together, inform the increased uptake of amino acids necessary for cell growth. Akt acts as the negative regulator of the two tumor suppressors, TSC-1 and TSC-2 (tuberin and hamartin of the tuberous sclerosis complex), by degrading them, which negatively regulates target of rapamycin (TOR), the master regulator of protein translation in mammalian cells. When Akt is spontaneously activated, it activates TOR by at least three independent mechanisms by:
- increasing ATP/ADP ratio by boosting glucose metabolism
- increasing saturated lipids to activate the lipid binding domain of TOR
- turning off the negative repressors of TOR, the most important mechanism of the three to fuel the cell with amino acids
This unique signaling pathway, which probably evolved from insulin-like growth factors, has become specialized and controls the perfect milieu for cell growth. To summarize, this pathway results in glucose uptake, fueling glucose through mitochondria to produce lipids, and amino acid uptake and use of those amino acids, through TOR-dependent regulation of translation, to increase protein synthesis. In all studies to date, this pathway is the genetic controller of cell size and growth.
Dr. Thompson believes this pathway is the fundamental and first transforming event that occurs in the vast majority of human tumors. However, although protein synthesis and lipids account for how cells get bigger, this pathway does not explain how the cells switch metabolism to replicate the genome.
The JAK-STAT is another critical metabolic pathway. It explains signal transduction control of ribosugar production and nucleotide biosynthesis, which is meditated by cytokine signal transduction through the Janus kinases (JAK) and STAT transcription factors. PI3K/AKT/TOR and JAK-STAT, together, regulate glycolytic metabolism. Activating mutations of either pathway make the cell absolutely dependent on glucose for continued ATP production and survival (because all metabolites and nutrients are diverted into synthetic processes).
How then might a cancer cell adapt its ability to use glucose if it has outgrown its blood supply? First, it is notable that glucose is likely to become a problem for such a cell before oxygen because glucose diffusion is via Brownian motion, whereas oxygen diffusion is based on partial pressures of the gas.
It is known that decreasing mitochondrial ATP/ADP ratio causes AMP production by adenylate kinase to restore ATP and also activation of AMPK. The tumor suppressor gene, LKB, activates AMPK in response to bioenergetic stress of cells. Non-transformed cells facing declining glucose concentration react by precisely arresting cell proliferation at the point where they must replicate DNA (G1-S boundary of the cell cycle), which is regulated by AMPK as a sensor. The surprise in these studies is that cell-cycle arrest is absolutely depended on the tumor suppressor p53; p53 null (-/-) cells cannot arrest the cell cycle when AMPK is activated, even when facing nutrient conditions that would impair their ability to go through S phase effectively. These cells continue to rumble through the G1-S checkpoint of the cell cycle and attempt to replicate their DNA in the face of limiting glucose concentrations.
This finding argues that normal non-transformed cells use p53 to respond to stress and ensure survival. One survival mechanism is AMPK activation of autophagy through p53 mediation, which promotes survival. Autophagy allows cells, in the absence of glucose uptake, to catabolize their intracellular contents in an orderly fashion to maintain ATP production by using them to fuel the mitochondria. Autophagy allows the non-transformed cell to adapt metabolically and is initiated by AMPK and dependent on LKB and p53.
Rapamycin, AICAR, and metformin activate autophagy as a cell survival strategy and can be used in cancer therapies. These findings were demonstrated in a series of experiments using Bert Vogelstein's HCT116 colon cancer cells that were isogeneic, but selected for p53 + or p53-. Animals were injected with tumor cells growing in response to activation of AMPK. In the absence of treatment, both tumors establish themselves, as expected, but p53-/- cells grew larger and more rapidly. When treated with AICAR or metformin, p53-/- cells were profoundly impaired, growing to less than half the size of p53+/+ wild-type (wt) cells. This finding suggests that tumor suppressors have a role in regulating intracellular metabolic adaptation under conditions of bioenergetic stress.
Hypoxia is another example of a bioenergetic stressor and has been studied in tumor cell growth and susceptibility to chemotherapy. Under conditions of hypoxia, HIF-1 acts a transcription factor to increase anaerobic metabolism. HIF-1 becomes stabilized, glycolytic target genes become activated, high throughput anaerobic glycolysis is induced, and ATP is formed, leading to survival. In findings to be published shortly, Dr. Thompson's group showed that this ability to induce and regulate HIF as a mediator of metabolism is entirely dependent on growth factor signal transduction. In response to hypoxia, HIF-1 acts to redirect growth factor-directed glucose uptake into the anaerobic pathway. HIF plays no role in glycolytic regulation in the absence of growth factors. The molecular explanation of this phenomenon is that HIF-1alpha is absolutely dependent on PI3K/AKT signal transduction for its ability to be transcribed at the RNA level or to be translated to be inducible by hypoxia.
These findings lead to the model in which the vast majority of tumor suppressors are intracellular adaptors of intermediate metabolism to fuel cellular responses to stress in bioenergetic or biosynthetic pathways. They are modulators of the cellular response to growth factor-derived signals that direct nutrient uptake. Predictions and therapeutic implications are listed at the beginning of this article.
Search Suggestions for Subscribers:
Numerous agents are currently in development, both preclinically and clinically, targeting several of the molecular markers mentioned in this article.
- For detailed reports on these agents: In the query screen for the New Drugs module, select the Targets report format then enter the marker name in the Target query box.
- To find more about the molecular markers themselves: In the Targets in Oncology module, enter the name of the marker in the Marker Designation query box.