DALLAS - A toxin isolated from a tiny, reclusive sea creature shrinks cancerous tumors in mice, Dallas scientists have found, without some nasty side effects caused by standard chemotherapy, such as significant weight loss or the loss of white blood cells.
The synthetic form of the toxin seems to work in an entirely new way from current cancer drugs - blocking the reproduction of cancer cells, apparently without affecting healthy cells, at least in mice.
The researchers say that if upcoming safety tests show people can tolerate the drug, they can investigate whether the drug also attacks tumors in patients. The best possible outcome, they say, would be a stronger cancer drug that's easier to stomach than today's chemotherapies.
The Dallas scientists say the compound shows promise as a cancer treatment - but they remain cautious.
Premium content for only $0.99
For the most comprehensive local coverage, subscribe today.
"Until this compound goes further into the development phases, we don't know whether it'll become a drug," said Patrick Harran, a chemist at UT Southwestern Medical Center who helped lead the new research. "I want to underpromise and overdeliver."
A research paper describing the compound's tumor-shrinking powers and a second paper on how the compound works appeared online Monday in the Proceedings of the National Academy of Sciences.
The journey to the compound's discovery began in the 1990s, when a team of researchers isolated it from a doughnut-shaped, soft-bodied sea creature offshore of the Philippines.
The creature, Diazona angulata, resembles a jellyfish in texture and lives in colonies anchored to rocks. Scientists believe it uses the toxin to repel predators. And researchers soon discovered the toxin - named diazonamide A - could kill cancer cells growing in lab dishes.
Scientists proposed an initial chemical structure for the compound in 1991. But Harran and colleagues showed it was incorrect. They reported the toxin's correct chemical structure in 2001. Since then, he and other UT Southwestern researchers have been trying to figure out how the toxin kills cancer cells.
In one of the new research papers, UT Southwestern scientists report that a derivative of the toxin - like many chemotherapy drugs - interrupts the division of cancer cells.
Traditional chemotherapy drugs block cell division by attacking microscopic strands that normally help pull apart a cell's chromosomes during division. If the cells can't divide, they die. The problem is, the traditional drugs can also damage healthy cells.
The new compound appears to work in an entirely new way, Harran said. The new toxin attaches to an enzyme associated with mitochondria, the tiny power plants that supply a cell with energy.
The Dallas researchers still don't understand why that should block the division of cancer cells, but it does.
Even more interesting is that, as far as early tests indicate, the drug doesn't appear to affect normal cells, Harran said. It may be that the enzyme targeted by the toxin is used only for the amped-up division of cancer cells, the researchers wrote.
It's also possible that the drug could affect healthy human cells, but not mouse cells, said UT Southwestern biologist Noelle Williams.
To test the compound's tumor-shrinking ability, researchers injected mice with human prostate, colon and breast cancer cells. Tumors formed in the mice but shrank when the mice were medicated with the compound. The test compound worked just as well as two traditional chemotherapy drugs, paclitaxel and vinblastine.
But while those drugs caused side effects familiar to cancer patients, such as blood cell problems and weight loss, the new compound did not. That's probably because the test compound kills cells in a different way than traditional chemotherapy, Harran said.
So far, the test compound looks quite promising, said Dr. Samuel Denmeade, a medical oncologist at the Johns Hopkins Kimmel Cancer Center in Baltimore who also investigates natural compounds as anti-cancer drugs. He was not involved in the UT Southwestern research.
"The data look as good as you'd want them to" at this point, Denmeade said.
But there's a long way to go. "This is like a baby lifting its head for the first time," he said. "It's not even crawling yet."