To do this, Lin fuses three genes in sequence: an easy to visualize tag, a protease from hepatitis C virus, and a synaptic protein of interest. The hepatitis viral protease likes to cut off anything attached to its own ends, so, left alone, the triple protein splits itself into its three constituents, leaving the synaptic protein untagged. If Lin administers an anti-hepatitis C drug that cripples the protease, all the fusion protein molecules made after that point remain intact with their tags still attached. The work remains at an early stage, but Lin has used the system to create transgenic flies, labeling a protein that's known to be a major structural component of synapses. Neither he nor Tsien thinks they've found the perfect marker for memory formation, but, says Lin, “We have a good enough candidate to begin work.”

Their “dream experiment” would be to find a master protein that is freshly made when a new synapse is born but is not being continuously refreshed in existing synapses. They would then create an animal with that protein labeled and train it in some task while giving it the anti-hepatitis C drug. All copies of the master protein formed from that point forward would be labeled. By observing where those proteins accumulate, they could see where nerve synapses had formed or expanded. From a practical standpoint this process is a bit more daunting than it sounds. They still don't know all the proteins that help create synapses. Tsien hopes other scientists will take up the memory question as well, so that they can divvy up the many potential protein candidates.

The effort that Tsien calls “foolish,” but is taking most of his energy, is his foray into cancer. It reflects a wish to do something about the disease that claimed the lives of his father and his Ph.D. supervisor.

When surgeons remove a tumor, they rely on their knowledge of how cancer tissues look and feel to remove as much as they can without taking out too much surrounding tissue. “The way we try to tell whether we have gotten it all is to take samples of what's left and send them to the pathology lab,” says Quyen Nguyen, a head and neck surgeon at the UC San Diego School of Medicine who also works in Tsien's lab. Examining tissue for cancer can take 10 or 20 minutes per sample, so that process can mean keeping a patient under anesthesia for an extra hour or longer.

Fluorescent proteins don't help here, because their main advantage is that they can be delivered by gene transfer, which is fine for experimental animals but not human patients. Nguyen and Tsien have found a way to load tumors with synthetic nanoparticles that not only are visible by magnetic resonance imaging but also glow on the operating table, making the edge of the tumor visible to the surgeon. The technology holds promise as a way to detect tumors and improve the success of surgery and, possibly someday, as a way to deliver drug treatment.

These nanoparticles are the molecular equivalents of self-adhesive name tags that come with a nonstick backing paper. As long as the “backing paper” is covering the “adhesive,” the molecules don't stick to cells. Malignant tumor cells secrete specific proteases that they use to chew their way through normal tissues to reach the blood and distant organs. The “backing paper” is carefully designed to fall off when chewed by the tumor cells' proteases, thus exposing the “adhesive” and coating the tumor cells with the nanoparticles.

	PAGE 1 2 3 4 5 6	Go Back | Continue