IN THIS ERA OF CANCER GENETICS, we’ve come to think of cancer as something like a car. Runaway cell proliferation is driven by accelerators, or oncogenes, that are stuck on go—in collusion with broken brakes, or tumor suppressors, that can’t control a tumor’s pedal-to-the-metal growth. A lengthening procession of targeted therapies (Herceptin, Gleevec, Tarceva, Zelboraf) attempt to block various accelerators, and they often succeed in bringing a tumor to a screeching halt—until the malignancy activates a different gas pedal and returns to the road.

Lately, increasing numbers of researchers have turned their attention to the excessive amount of fuel that accelerating cancers require, and to how tumors use the fuel. That focus on the metabolism of cancer cells bridges the fields of oncology and cellular metabolism, says Lewis Cantley, a biochemist and the director of the Cancer Center at Beth Israel Deaconess Medical Center, whose work has helped resurrect this long-sidelined approach to fighting tumors. “If we can understand the unique ways that cancer cells use energy, then we may be able to exploit their weaknesses,” Cantley says.

Metabolism, in terms of the whole body, is a balancing act. It’s part of a calories in, energy out equation, the reason that eating too much and exercising too little leads to weight gain. But metabolism also happens at a cellular level. Normal cells take in sugar, or glucose, from the blood to generate heat and to fulfill their appointed tasks. Cancer cells do the same thing, but they burn through much more glucose—a supercharged uptake that’s visible on FDG (fludeoxyglucose) PET scans. Oncologists commonly use this imaging technique to help locate tumors, establish the stage of those cancers’ progression, and determine whether they have stopped growing in response to therapy or are relapsing or metastasizing.

Many researchers now believe, however, that taking advantage of cancer’s appetite for glucose could go beyond diagnosis and lead to effective treatments. It’s impossible to deprive a patient’s cancer cells of glucose, because the liver will synthesize additional sugar if the bloodstream has an insufficient supply. Yet it might be feasible to scramble the distinctive way that cancer cells metabolize glucose and starve hungry cancer cells to death—or at least to use that approach to complement other therapies.

DECIPHERING CANCER METABOLISM BEGINS WITH adenosine triphosphate, or ATP—the energy currency that all cells use to power what they do. Under most conditions, cells use oxygen to burn glucose in their mitochondria, located in cells’ cytoplasm, the gelatinous filling around the nucleus. For this, the cells use the tricarboxylic acid (TCA) cycle, also known as the Krebs cycle and as oxidative phosphorylation. In the TCA cycle, glucose, oxygen and other chemicals are modified, combined or broken down into components (metabolites), some of which re-enter the cycle while others go off to perform other duties. This process generates ATP, with carbon dioxide as a by-product. Occasionally, though, a cell will become oxygen-depleted, or anaerobic, such as when someone tries to sprint 500 yards. “For the last 100 yards, your leg muscle cells aren’t getting oxygen,” says Raul Mostoslavsky, a molecular biologist at Massachusetts General Hospital Cancer Center. Muscle cells are then forced to switch to glycolysis. Compared with normal cellular metabolism, which can generate 36 units of ATP per molecule of glucose, glycolysis seems grossly inefficient, producing just two units, with lactate as a by-product. Lactate is acidic, and it’s what makes muscles ache following anaerobic exercise, says Mostoslavsky.

But extreme exercise isn’t required to force cancer cells to opt for glycolysis. In 1924, Otto Warburg, a German biochemist who later won the Nobel Prize, observed that malignant cells use glycolysis even when they have enough oxygen. Aerobic glycolysis, now known to happen in almost all cancers, is called the Warburg effect.

Why would cancer cells switch from a mechanism that produces maximum energy to such a wasteful use of glucose? Warburg proposed that it resulted from damaged mitochondria, leaving the cell no other choice. He also contended that this altered metabolism was the cause of cancer, and for decades that notion was at the center of cancer research.

Cantley, as a graduate student in biochemistry during the 1970s, was part of that wave. But researchers eventually learned that the mitochondria in most cancer cells are not in fact damaged, and that most cancer cells freely select glycolysis. And once James Watson and Francis Crick had defined the structure of DNA, genetic research blossomed. The first cancer-causing oncogenes were discovered, and the focus of efforts to understand and control cancer shifted to its genetic underpinnings. Cancer metabolism, it seemed then, was a mere side effect of what happened in genes.

That view persisted until recently, when the conventional wisdom began to reverse itself again. “It took about 20 years to figure out what oncogenes do—and guess what?” says Cantley. “They reprogram the metabolism, and that promotes cancer. If we turn off one oncogene with a targeted therapy, another one turns on to rewire some other metabolic pathway. So why not just target metabolism?”

UNTIL ABOUT FIVE YEARS AGO, THOUGH, science couldn’t explain why cancer cells with sufficient oxygen would consistently opt for glycolysis—which, after all, produces just a fraction of the energy that normal metabolism can generate. Then Craig Thompson, a medical oncologist and president and CEO of Memorial Sloan-Kettering Cancer Center in New York City (along with Cantley and one of his postdoctoral students, Matthew Vander Heiden) rethought the problem. What if there was something else the cells needed even more than energy? At the time, Thompson was director of the Abramson Cancer Center at the University of Pennsylvania, and as a medical oncologist with training in immunology, he was studying normal, healthy white blood cells, or lymphocytes, to understand how fast-dividing cells make the decision to divide. Rapidly proliferating lymphocytes also choose glycolysis, and he realized that it was because glycolysis gives those cells more of what they need to replicate themselves: the building blocks of new cells.

When cells are growing fast or dividing, they have to create new proteins for receptors, new lipids for membranes, new nucleotides for DNA and new fatty acids for cholesterol. In glycolysis, more of the components of glucose are available to form the bricks and mortar of growing and dividing cells. Thanks to that insight, subsequently confirmed in numerous labs, “the whole field exploded,” says Mostoslavsky.

This new view of cancer metabolism also conjures up a new metaphor: a cabin in the woods. Trees (glucose) can be burned as fuel for heat, or they can be used to build a cabin. A normal cell is like a cabin that is content to burn a few logs at a time for warmth. A cancer cell is a cabin that wants to expand (cell growth) and then build a cluster of additional cabins to create a tumor (cell division). In biochemistry, the heat-generating metabolism, called catabolism, happens in the mitochondria, in the TCA cycle. The building-block-producing metabolism, or anabolism, uses glycolysis. “Warburg thought glycolysis was wasteful,” says Thompson. “It’s not. It allows the cell to deal with its metabolic problem of how to grow and multiply.”

In most adult tissues, cells go about their daily business—absent a command from a growth factor, such as epidermal growth factor (EGF), they don’t really need to grow, much less divide. And even when a growth factor sends its signal to grow by attaching to receptors on the cell surface, a cell must decide whether it has sufficient nutrients to obey.

That work is carried out through a molecular pathway known as PI3K—phosphoinositide 3 kinase—an enzyme Cantley discovered during the 1980s that relays signals and activates other enzymes or genes. The PI3K signaling pathway is normally tightly controlled, but in many cancers that control is absent. Any of several mutations and aberrations can push the pathway into a hyperactivated state, pressing a cell to grow and divide more than its environment calls for. The pathway increases the activation of HIF-1a (hypoxia inducible factor 1 alpha) and produces more glucose transporters on the cell surface—both key changes that promote glycolysis in cancer cells.

Insulin, the hormone that normally tells muscle and fat cells to take in glucose, also activates PI3K, Cantley discovered. Those cells then use the glucose for anabolic growth. It turns out that lots of epithelial tumors (carcinomas) have insulin receptors too. “The cancer cells are getting what muscle and fat cells usually get,” he says, “and that changes how they function”—in ways that fulfill cancer’s need for new building blocks.


RESEARCHERS NOW KNOW THAT A CANCER CELL derives benefits from glycolysis that go beyond having plenty of raw materials for new cells. For one thing, glycolysis can still produce a lot of energy. Though the TCA cycle generates many more units of ATP per molecule of glucose, glycolysis happens so quickly that it actually generates more ATP per second than the normal process does. Moreover, cell metabolism is not an either-or proposition. Cancer cells using glycolysis also continue to metabolize glucose in their mitochondria using the TCA cycle, to produce still more energy. That, in turn, generates free radicals—unstable molecules that can damage DNA, creating additional mutations that allow the tumor to become more invasive, infiltrating tissues and migrating to other parts of the body. Glycolysis’ waste product, lactic acid, might also make surrounding tissue more vulnerable to a tumor’s infiltration.

Moreover, cancer’s metabolic reprogramming goes beyond glycolysis and the Warburg effect. Malignant cells also depend on another fuel, glutamine, an amino acid that cells can metabolize to produce both energy and building blocks. Glutamine, unlike glucose, contains nitrogen, a key component of the nucleotides in DNA, RNA and ribosomes, which make up much of the mass of a cell—and that all have to be duplicated when cells divide. It turns out that the oncogene Myc, essential for glycolysis, also directs cancer cells to scavenge glutamine out of the bloodstream. By inducing both glycolysis and glutamine metabolism, Myc uniquely supports the synthesis of the rapidly proliferating cells.

Not all cancers show up on FDG PET scans, and that could be because some use glutamine metabolism rather than glycolysis. Or they could depend on still another nutrient, the amino acid glycine. A May 2012 article in Science found that, in a study of 1,300 samples of tumors from early-stage breast cancer patients, those whose tumors had higher levels of glycine synthesis were more likely to die from the disease.

Researchers know very little about how the body regulates glycine metabolism. Yet its contribution to tumor cell proliferation only increases the evidence that changes in metabolism are a cause of cancer and not just a consequence, according to Leif Ellisen, a cancer genetics researcher and oncologist who directs the MGH Translational Research Laboratory. “Knowing that these metabolic changes actually drive cancer growth gives us an important rationale to target their pathways as therapy,” he says.

That case became even stronger with the discovery in 2008 of a mutated gene for a metabolic enzyme called IDH (isocitrate dehydrogenase) in some leukemias, brain cancers, and gall bladder and gastrointestinal cancers. The mutation allows cells to synthesize an “oncometabolite” called 2-hydroxyglutarate (2HG), which modifies other enzymes, promotes the production of lipids and enhances tumor growth. IDH is mutated early in the development of those cancers, supporting the theory that cancer cells evolve toward altered metabolism, “selecting” mutated genes as they develop. Now researchers are finding additional mutated metabolic enzymes that can behave as oncogenes, says Chi Dang, a medical oncologist and cancer biologist who directs the Abramson Cancer Center, and such discoveries are helping convince skeptics that alterations in metabolism are not mere “epiphenomena” or side effects.

THESE DISCOVERIES SUPPORT THE NOTION THAT by rewiring their metabolism, cancer cells become “addicted” to nutrients—a weakness ripe to be exploited. When normal cells sense a shortage of nutrients, they can slow down to conserve their energy. But a cancer cell’s imperative is to grow and divide, and if it doesn’t get sufficient nutrients, it may go into autophagy—consuming itself in an attempt to produce the building blocks for new cells. Autophagy eventually kills cells, so new therapies that deprive cells of nutrients or otherwise thwart metabolism will likely cause more harm to cancer cells than to normal cells. In 2007, Lewis Cantley, Craig Thompson and Tak Wah Mak, a researcher at the University of Toronto, founded Agios, a Boston biotechnology company, to develop anti-metabolism agents, and many other biotech firms and pharmaceutical companies are exploring similar strategies.

Much of the effort to impede cancer metabolism focuses on the PI3K pathway that is hyperactivated in so many cancers. PI3K inhibitors are the subject of some 100 clinical trials, including one that will test whether a cancer treatment’s early failure to reduce glucose uptake, as measured by FDG PET, can predict whether the therapy will fail to shrink a patient’s tumor. Most researchers expect that combining such agents (and other anti-metabolism drugs) with targeted therapies or chemotherapy will work better than any single drug can do on its own.

Researchers also find the pyruvate kinase (PK) enzyme intriguing. PK is a gatekeeper that helps determine whether a cell metabolizes glucose using the TCA cycle or diverts it to other pathways for anabolic growth. PK comes in several forms, and many normal adult cells use PKM1, which favors the TCA cycle. When cells become cancerous they use PKM2 in glycolysis instead, and Myc and HIF-1a, the oncogenes that affect metabolism, may influence that switch. “Because PKM2 is a common denominator in all cancers and clearly distinguishes them from many normal cells, it is a potential drug target,” says Matthew Vander Heiden, now a cancer researcher at the MIT Koch Institute for Integrative Cancer Research who studies this enzyme.

Another possibility is to target glutamine metabolism. According to John Blenis, a cell biologist and biochemist at Harvard Medical School, many cancer cells in the lab have difficulty surviving without glutamine, whereas normal cells in culture can cope with low glutamine by shutting down growth and energy use. As with glucose, the body makes more glutamine if there’s a shortfall, so it’s impossible to deplete directly. However, in 2010, Blenis, with Vander Heiden, Cantley and colleagues, found that interfering with how cells metabolize glutamine can limit tumor cell growth and survival.

Yet cancer metabolism, like everything else that happens in tumors, is very complex, and a therapy that’s effective against one metabolic process may not work against another. Still, laying siege to cancer cells’ supply lines may cast a wider net than targeted therapies aimed at rare oncogenic mutations. Indeed, one widely used cancer treatment—methotrexate—works by interfering with the metabolism of folate, an anabolic brick used to build new DNA. And drugs that treat other diseases by affecting metabolic pathways might also turn out to inhibit cancer.

SO WHERE DOES ALTERED METABOLISM RANK as a cause of cancer? “There’s quite a debate about this in the community now,” says Daniel Haber, director of the MGH Cancer Center. He notes that the altered metabolite produced by the IDH mutation appears to indirectly affect how genes are regulated. It does so by inhibiting other enzymes that modify chromatin, the protein packaging around spools of DNA, by opening or closing a spool to raise or lower the activity level of a gene in that area. Thus, Haber muses, altered cancer metabolism is not just about energy supply, but also directly affects gene expression through regulating chromatin, and that “closes the loop” between metabolism and genetics—but still poses the chicken-or-the-egg problem of which came first.

One way of thinking about that problem, Ellisen suggests, is that “a cell needs successive mutations, or hits, to become a cancer cell.” Some of those hits affect tyrosine kinases, enzymes that promote cancer by stimulating abnormal cell growth and proliferation. “But those activities require energy and building blocks,” he says. “So a cell that rewires its metabolism early on to scarf up nutrients from the environment is more likely to do what the tyrosine kinase pathway sets in motion; it’s more likely to have all the things it needs to become a tumor.”

And some metabolic mutations, which themselves can promote cancer-causing changes in the cell, will be more successful if they’re followed by mutations in tyrosine kinases or transcription factors that stimulate cell growth. “Maybe the normal versions of oncogenes and tumor suppressors originally evolved to regulate metabolism, and their mutations alter metabolism in ways that inevitably become oncogenic,” says Thompson.

“It took us 80 years to bring the revealing observations of Otto Warburg back from obscurity,” says Mostoslavsky at MGH. “Now that we widely acknowledge cancer metabolism as a main player in cancer, it will take much less time to exploit metabolism in tackling this devastating disease.”