Researchers at Stanford University have discovered that they may be able to tweak cancer cells so that they summon their own demise.

One of the biggest hurdles in developing new cancer drugs comes from identifying extremely precise targets so the drugs kill only cancer cells and spare those that are healthy. Now, researchers at Stanford University have discovered that they may be able to tweak cancer cells so that they summon their own demise.

The new research shows that the protein, calreticulin, is present on a wide variety of cancer cells but rarely on healthy cells. Calreticulin is viewed as an attack signal by white blood cells known as macrophages. Some cancer cells have evolved ways to suppress calreticulin and evade the immune system by restraining the macrophage attack. However, the new finding, published December 22, 2010, in the journal Science Translational Medicine, suggests that a drug under development might be able to remove this restraint and permit macrophages to destroy cancer cells while sparing virtually all other cells in the body. The paper’s first author is Stanford University medical student Mark Chao, who did the work while participating in HHMI’s Medical Research Fellows Program. The program, which is currently funded by a $2.5 million HHMI initiative, provides a year of biomedical research training to select medical, veterinary, and dental students.

Only the cancer cells were eaten by macrophages. The normal cells were spared.

Mark Chao

Chao’s research is part of an ambitious preclinical cancer drug development program in the Stanford laboratories of Ravindra Majeti and Irving Weissman. In a series of papers since 2009, the researchers have shown that macrophages, which normally gobble up infected or otherwise damaged cells, will also feed on a variety of cancer cells, something they have demonstrated in mice and in cell culture systems. But macrophages will only attack cancer cells if CD47— a protein commonly expressed on the surfaces of the cancer cells—is blocked with an antibody.

The research teams concluded early on that CD47 tells macrophages not to eat certain cells. By removing that “don’t eat me” signal, the researchers found that macrophages unleashed their attack on cancer cells. But CD47 is found also on many non-cancerous cells, so the anti-CD47 antibody they used in the experiments coated both normal cells and cancer cells. In principle, that would tell the macrophages to eat both cell types. “Yet only the cancer cells were eaten by macrophages,” Chao says. “The normal cells were spared, and when we gave the antibody to healthy mice, it caused minimal toxicity.”

In the experiments published in Science Translational Medicine, Chao and his colleagues set out to determine why the normal cells were protected. They thought that another signal must be involved, perhaps a macrophage-attack signal that appears only on the cancer cells. High on their list of suspects was calreticulin, a “kill me” protein already known to appear on certain cells after DNA damage and other kinds of stress. They also knew that calreticulin binds to a receptor on macrophages, causing the macrophages to eat the damaged cells. Could it work the same way in cancer cells?

The key test, according to Chao, was whether the macrophages behaved differently when they interacted with cells in the presence or absence of calreticulin. When the researchers incubated both macrophages and cancer cells with anti-CD47 antibodies, the macrophages ate the cancer cells, as expected. Then the scientists blocked calreticulin expression in the cancer cells, and the cells survived. The team also found that increasing calreticulin expression on cancer cells correlated with higher expression of CD47, suggesting that increasing CD47 expression might be a trick that tumors evolve to avoid being destroyed by macrophages.

The finding seems promising because appearance of both calreticulin and CD47 is not limited to one or a few cancer cell types. “This system is the first that is universal to all cancers, as far as we can tell,” Weissman says. With funding from the California Institute of Regenerative Medicine and the UK’s Medical Research Council, the Majeti and Weissman labs are developing their anti-CD47 antibody treatment for initial testing against acute myelogenous leukemia, although they expect it to be effective against many and perhaps all cancers. “This finding about calreticulin tells us that we can go forward and target CD47 and expect that we won’t get much toxicity, if we’re careful,” Weissman explains.

The researchers are now attempting to solve one of the remaining mysteries about calreticulin, namely why it appears on cancer cells in the first place. “It might be that it appears as a byproduct of cellular damage, and normally triggers the clearance of incipient cancer cells by macrophages, except when those cells happen to express higher levels of CD47,” Chao says. Calreticulin might also actively drive cancer growth.

Chao and his colleagues have looked at the levels of calreticulin expression on a variety of human cancers and found that higher calreticulin expression correlated strongly with worse clinical outcomes. “There seems to be an advantage for the cancer cells that have this higher calreticulin expression, although as yet we don’t know what that advantage is,” he says.

Chao became an HHMI Medical Research Fellow in 2007, after witnessing first-hand the limits of current oncology drugs. “I was seeing leukemia patients (in the clinic), and it was clear that they needed therapies that target cancer more effectively without harming normal cells,” which is common with current treatments like chemotherapy and radiation, he says. Since joining the HHMI program and the Majeti and Weissman labs, he has completed a Ph.D. program in cancer biology at Stanford and plans to spend his career doing translational research. “Without the chance to work in a lab, I might have gone in a completely different direction,” he says.

For More Information

Jim Keeley 301.215.8858