To avoid being destroyed by our immune systems, cancer cells engage in a bit of trickery. As they divide to form tumors, they fly under the radar of macrophages, immune cells whose job it is to ingest dead cells and dangerous invaders. Today, many cancer patients are treated with antibody drugs that work in part by marking tumor cells for destruction by macrophages. Although these drugs have extended lives, they don't always work very well—partly because cancer cells fight back by sending a "don't eat me" signal to the immune cells. Now, researchers have created a small protein that shuts off this signal and dramatically boosts the power of antibody drugs to shrink tumors in mice.
Stem cell biologist Irving Weissman of Stanford University in Palo Alto, California, has been studying a protein responsible for broadcasting this "don't eat me" signal. Known as CD47, it protects leukemia cells and other cancer cells from rampaging macrophages. Last year, Weissman's group reported that an antibody that blocks CD47 on the cancer cells can spur macrophages to destroy tumors in mice. The anti-CD47 antibody will soon be tested for safety in humans with $20 million in funding from the California Institute for Regenerative Medicine.
As promising as this approach is, antibodies have drawbacks—their relatively large size limits how easily they can penetrate tumors, for example, and they can have toxic side effects—so Weissman's group has also teamed up with structural biologist Christopher Garcia's lab at Stanford to test a different way of blocking CD47. Garcia's team and Weissman describe the new work today online in Science.
The collaborators began by studying a protein called SIRPα that pokes out from the surface of macrophages and connects with CD47 to receive the tumor cell's "don't eat me" signal. Their idea was to block CD47 with a free-floating, synthetic form of SIRPα that was engineered so that the macrophage's real SIRPα couldn't latch onto the cancer cell and be tricked into leaving it alone. Garcia's lab synthesized many versions of the SIRPα protein, determined their structures, and found that two forms bound 50,000 times more tightly to CD47 than the natural SIRPα receptor.
Added to a petri dish of cancer cells and macrophages, the synthetic SIRPα proteins didn't do much—the macrophages still ignored the cancer cells. However, when the researchers threw in a tumor-specific antibody drug, which they realized was needed to draw the macrophages' attention to the cancer cells in the first place, the combination packed a powerful punch in cell cultures as well in mice implanted with cancer cells. For example, while lymphoma tumors merely grew more slowly in mice given either the drug rituximab or SIRPα proteins, the tumors virtually disappeared for at least 7 months in most mice treated with both drugs. And adding SIRPα to the breast cancer drug Herceptin shrank tumors faster in mice with the disease. "SIRPα weakens the cancer cell's ability to protect itself from destruction," Garcia explains.
Because the synthetic SIRPα proteins are relatively nontoxic, Garcia and Weissman hope that they can find funding to quickly develop them into an experimental drug. "I would like to see this tested in humans as soon as possible," Garcia says.
Cancer immunologist Drew Pardoll of Johns Hopkins University in Baltimore, Maryland, says that the idea of blocking CD47 with a small protein instead of a large antibody is "pretty nifty. I'm not aware that this sort of approach has been taken before." However, he cautions that as with other new treatments that work by lifting the brakes on the immune system, there could be side effects. "These negative regulatory pathways aren't there for no reason at all" but can have important roles for normal cells, he says. For example, the CD47 "don't eat me" signal also protects red blood cells from destruction by the immune system.
University of Amsterdam cell biologist Timo van den Berg adds that while "the data are quite convincing and exciting," because CD47 is expressed on most cells in the body, these cells will suck up the SIRPα protein, which could make it difficult to get enough to the tumor cells. His own group is working on an antibody that blocks the SIRPα receptor, which appears mainly on macrophages, as a way of preventing these immune cells from receiving the cancer cell's deceptive message. Garcia responds that although he thinks his team's approach will work well in patients, "certainly blocking from the other direction is a viable strategy and worth looking at."