Novel Cancer Hypothesis Suggests Antioxidants Are Harmful

A new hypothesis that focuses on reactive oxygen species (ROS) proposes that antioxidant levels within cancer cells are a problem and are responsible for resistance to treatment.

The theory destroys any reason for taking antioxidative nutritional supplements, because they "more likely cause than prevent cancer," according to Nobel laureate James Watson, PhD, from Cold Spring Harbor Laboratory, New York.

Dr. Watson, who shared the Nobel prize for unraveling the structure of DNA, regards this theory as being "among my most important work since the double helix," notes a press release from his institution, where he has been director since 1968.

The theory was published online January 8 in Open Biology.

Dr. Watson explains that the vast majority of agents used to directly kill cancer cells, including ionizing radiation, most chemotherapeutic agents, and some targeted therapies, work by generating — either directly or indirectly — ROS that block key steps in the cell cycle.

This generation of ROS creates a hypoxic environment in which cancers cells undergo a transformation from epithelial to mesenchymal cells (EMT).

These transformed cells almost inevitably posses very high amounts of antioxidants, which effectively block the effects of anticancer treatments, Dr. Watson notes. Once a cancer becomes resistant to chemotherapy, it usually is equally resistant to ionizing radiation, he points out.

In addition, these transformed EMT cancer cells generate free-floating mesenchymal cells, which have the flexibility and movement that allows them to metastasize to other body locations (brain, liver, lung). "Only when they have moved do most cancers become life-threatening," Dr. Watson notes.

Interestingly, the widely used antidiabetic drug metformin has been shown to preferentially kill mesenchymal stem cells. "In a still much unappreciated article published 3 years ago," metformin added to chemotherapy "induced prolonged remission if not real cures" in mouse models of cancer (Cancer Res. 2009;69:7507-7511), Dr. Watson writes. He notes that clinical trials are currently looking to see if adding metformin to chemotherapy provides clinical benefits, but adds that diabetics who have been using metformin regularly have a reduced incidence of many cancers.

Resistance to Therapy From Antioxidants in Cancer Cells

Dr. Watson proposes that anticancer therapies work by generating ROS, which cause apoptosis. However, as cancer cells evolve, they produce antioxidant proteins that block this effect, such as glutathione, superoxide dismutase, catalase, and thioredoxin.

The fact that cancer cells largely driven by RAS and Myc are among the most difficult of cancers to treat could be due to their high levels of ROS-destroying antioxidants, Dr. Watson argues. High antioxidative levels might also explain the effective incurability of pancreatic cancer, he adds.

If this theory is correct, then drugs that lower antioxidant levels within cancer cells would be therapeutic. In fact, the ROS-generating agent arsenic trioxide has been shown to reduce levels of glutathione and thioredoxin. Arsenic trioxide is currently being used to treat promyeloblastic leukemia, but this theory suggests that the drug could be useful in many major cancers.

Nutritional Antioxidants Could Be Harmful

One far-reaching implication of this theory is that antioxidants as nutritional supplements, including beta-carotene, vitamins A, C, and E, and selenium, could be harmful in cancer.

For years, such supplements have been widely hyped for cancer prevention and/or treatment, as has encouragement to eat colorful fruit and berries, which contain antioxidants.

However, Dr. Watson warns that recent data strongly hint that much of the untreatability of late-stage cancer might be the result of "its possession of too many antioxidants, [so] the time has come to seriously ask whether antioxidant use more likely causes than prevents cancer."

Many nutritional intervention trials have shown no obvious effectiveness in preventing gastrointestinal cancer or in lengthening mortality, he writes. "In fact, they seem to slightly shorten the lives of those who take them."

Hence, he concludes, "blueberries best be eaten because they taste good, not because their consumption will lead to less cancer."

论文连接：http://rsob.royalsocietypublishing.org/content/3/1/120144.full