After the Duke RNA binds its target on the surface of prostate cancer cells, it is eventually dragged inside the cell. Once inside, the RNA is cleaved in two by a protein native to the cell, freeing the gene-silencing region to find and guide the destruction of its target. RNA interference leads to the destruction of the intermediary between DNA and proteins, called messenger RNA. The Duke therapy destroys the messenger for a gene whose protein prevents prostate cancer cells from dying, even when outside signals tell the cells to do so. With this protection removed, cancer cells died.

Sullenger says that in principle it is possible to use the all-RNA technique to design therapies for many different diseases and infections. Hundreds of tumor markers, for example, are known. Sullenger's lab is engaged in the trial-and-error process of finding RNA sequences that bind protein markers and has found many.

Other researchers have had success with designing carriers to take therapeutic RNA into specific cells using various methods, including antibodies to target selective cells. But Sullenger says his method has several advantages. Antibodies are bulky proteins that don't penetrate tissue as well as plain RNA. Another alternative is to bind the RNA to a cholesterol molecule, which works well for delivering the drug to the liver or kidney. These approaches are complicated: researchers make two molecules (the RNA and, for example, a protein), then attach them to each other.

"Instead of mixing and matching, we decided to take advantage of RNA's binding abilities," says Sullenger. The Duke RNA therapy can be made in one step, moves through tissue easily, and could in principle be designed to target any cell in the body, Sullenger says. "All cells are a sink for these RNA compounds" when they are not specifically targeted, says Sullenger, which means large amounts would have to be introduced to the body to achieve a therapeutic dose in, for example, a tumor in the kidney. This saturation dose could be expensive and could have toxic side effects.

Sullenger's group is now working to prove that the RNA therapy will avoid these pitfalls. For the prostate cancer experiments, they injected the RNA directly into the mice's tumors. They must now demonstrate that the therapy will still reach the tumors when given systemically, by injection into the blood. Sullenger says he would also like to monitor more carefully for immune reactions in the mice. And he hopes to test the RNA therapy in humans if all goes well with these further experiments.

So far, says Rossi of the Beckman institute, the Duke approach seems a "much more efficient" way to get RNA into cells and could greatly reduce the cost of RNA interference therapies. Says Alnylam's Greene: it has "great promise."