Growing tumors need new blood vessels. Cancer researchers have spent years working to starve tumors by blocking this blood vessel growth, or angiogenesis, with mixed success (Science, 22 March, p. 2198). Now a team has tackled the problem from a novel angle: They packed a tiny particle with a gene that forces blood vessel cells to self-destruct, then "mailed" the particle to blood vessels feeding tumors in mice.
In the mid-1990s, David Cheresh, a vascular biologist at the Scripps Research Institute in La Jolla, California, and others found signatures specific to various types of blood vessels that they termed "zip codes." One zip code belongs to a class of membrane proteins called integrins; it's apparently always present on angiogenic, or newly growing, blood vessels but rarely on established ones. The integrin, called ?v?3, has another quality that would turn out to be useful: It's expert at propelling viruses or other small particles into cells.
To target new blood vessels, Cheresh's team studded nanoparticles with molecules that bind to ?v?3. They also embedded in them mutant copies of a gene called Raf-1. Raf-1 moderates cascades of molecular signals that govern blood vessel growth. Once the particles reached tumor blood vessel cells, the researchers hoped that ?v?3 would propel the mutant gene inside the cell, interfering with angiogenesis and forcing the cell to self-destruct.
Indeed, when the researchers infused a dose of these particles into the mice that were injected earlier with malignant cells, they found that a single treatment erased large tumors--the equivalent of a 2-kg tumor in an 80-kg person--in about 6 days. Animals with metastases to the lungs or liver also saw most of their tumors disappear. In contrast, the team reports in the 28 June issue of Science, mutant mice given nanoparticles without the ?v?3 beacon died after a day or two.
"It is a very provocative paper, which I think will become a landmark in angiogenesis research," says antiangiogenesis pioneer Judah Folkman of Children's Hospital in Boston. Still, one still has to prove the treatment works, says Philippe Leboulch of Harvard Medical School in Boston. Many more animal tests must be conducted before the therapy can even be attempted in humans--and before biologists and patients will know whether it's effective.
David Cheresh's lab