By Janet Fang

Short sequences of RNA that can effectively turn off specific genes have for the first time been used to treat skin cancer in people.

The technique, called RNA interference (RNAi), gained its inventors a Nobel Prize in 2006, but researchers have struggled to get it to the clinic, partly because of problems in getting the molecules to their target.

Now, Mark Davis from the California Institute of Technology in Pasadena and his colleagues have found a way to deliver particles containing such sequences to patients with the skin cancer melanoma. When analyzing biopsies of the tumors after treatment, they found that the particles had inhibited expression of a key gene, called RRM2, needed for the cancer cells to multiply. Their research was published online on March 21 in Nature.

The researchers created the particles from two polymers plus a protein that binds to receptors on the surface of cancer cells and pieces of RNA called small-interfering RNA, or siRNA, designed to stop the RRM2 gene from being translated into protein. The siRNA works by sticking to the messenger RNA (mRNA) that carries the gene's code to the cell's protein-making machinery and ensuring that enzymes cut the mRNA at a specific spot.

When the components are mixed together in water, they assemble into particles about 70 nanometers in diameter. The researchers can then administer the nanoparticles into the bloodstream of patients, where the particles circulate until they encounter "leaky" blood vessels that supply the tumors with blood. The particles then pass through the vessels to the tumor, where they bind to the cell and are then absorbed.

Once inside the cell, the nanoparticles fall apart, releasing the siRNA. The other parts of the nanoparticle are so small, they pass out of the body in urine. "It sneaks in, evades the immune system, delivers the siRNA, and the disassembled components exit out," Davis says.

Fire at will

The study describes the science behind a phase I trial assessing the safety of the technique in 15 patients. When researchers analyzed tumor samples from three of the patients who volunteered samples, they found fragments of the mRNA in exactly the length and sequence they would expect from the design of their siRNA. And in at least one patient, the levels of the protein were lower than they were in samples of the tumors taken before treatment. They also found that patients who were given higher doses had higher levels of siRNA in their tumors. "The more we put in, the more ends up where they are supposed to be, in tumor cells," Davis says.

Researchers will need more data from clinical trials to ensure that such therapies are safe to use in people. But Davis says that his study means there is now direct evidence that nanoparticles and RNAi can be used to attack harmful genes in humans--and not just in the test tube. "What's so exciting is that virtually any gene can be targeted now," he says. "Every protein now is druggable."

Davis says that by targeting specific genes he hopes these treatments will not have major side effects. "My hope is to make tumors melt away while maintaining a high quality of life for the patients," he says. "We're moving another step closer to being able to do that now."

But some researchers are concerned that the treatment has not been tested on more patients and that more samples were not taken from each patient. Molecular biologist Thomas Tuschl from Rockefeller University in New York says it is "exciting that such nanoparticles in multiple dosing schemes can reach the tissue and apparently have measurable effects". That is, he says, if one wants to believe the data from the single patient who had lower levels of the protein. "I hope these findings can be confirmed in the future," he adds.

Biomedical engineer Daniel Anderson from Massachusetts Institute of Technology in Cambridge has also been trying to develop RNAi delivery systems, and he thinks the data are a great start. "Generally, people who worry about making therapeutics understand that animals are not people," he says. "There are a lot of really exciting data with animals, but ultimately, the usefulness of these types of drug-delivery systems must be evaluated in humans, and that's why this is an important study. But it's not like we're done."