Margaret Black designs weapons that make their targets self-destruct. She’s not a military strategist or explosives expert, though, but a molecular biologist working to perfect a way to trick cancer cells into killing themselves.

Her approach is called suicide gene therapy. It works by a sneaky route that even a fabled spy like Mata Hari could appreciate. In conventional chemotherapy, the patient is given a drug that kills any cells in the body that are replicating their DNA. Suicide gene therapy works by infiltrating cancer cells and getting them to make the drug that will do them in.

Best of all, says Black, suicide gene therapy can minimize the nasty side effects commonly suffered by patients on chemotherapy.

A patient on suicide gene therapy receives two substances: a prodrug, which by itself has little effect; and a suicide gene that codes for an enzyme that converts the prodrug into a toxic form of the drug.

Black, who is an associate professor in the Department of Pharmaceutical Sciences, says most gene therapists are working on how to target cancer cells specifically, so the suicide gene doesn’t end up in normal cells. Her research focuses on what happens once the gene and prodrug get into the cell. She’s trying to make a form of the gene that is more deadly for cancer cells and that produces less spillover of toxin from cancer cells to normal cells.

She does that by making mutations in the natural form of the enzyme’s gene and then screening for mutant enzymes that can work with just a tiny amount of prodrug.

“What we’re trying to do is evolution in a test tube,” says Black. “Evolution is generally something that happens one step at a time, one mutant at a time. What we’re trying to do is make leaps at a time by introducing many mutations simultaneously.”

One approach is to hit the gene with random mutations. That works, but it’s inefficient. Black prefers to target areas of the gene she thinks might shift the enzyme’s activity without destroying it altogether. Then she puts the mutant genes into bacterial cells and uses a two-step screening process to find the mutants that are most likely to succeed as suicide genes. First she identifies the mutants that still have a functional enzyme.

The second test is a bit tricky. The mutants she wants -those best able to turn the harmless prodrug into the deadly drug-will die during the test. Black keeps samples of each mutant alive and healthy in other containers until the test is completed. Those whose brethren die when fed the prodrug move on to more detailed analysis of their mutant enzyme.

The screening process looks reasonable on a diagram. Then you see the numbers she’s dealing with. In one series of experiments, Black and her students screened more than a million mutants. They found two that were good candidates as therapeutic agents.

“It’s a numbers game,” she shrugs. “There’s a point where you can go crazy doing this. My students will tell you that, because they’re in the process of screening a lot.”

In work recently reported in Science Magazine, Black and several colleagues at the Fred Hutchinson Cancer Research Center and the University of Washington described a new method to streamline the process. They used a special computer program to predict the effects of mutations at various points in the enzyme, and came up with a triple mutant that looked promising. Black is now testing the new weapon for its ability to kill cells from a rat brain tumor.

Black didn’t start out as an espionage agent in the cancer wars. Her main interest has been the enzymes involved in DNA synthesis-how they work, and how their structure relates to their function. The cancer connection came about in 1992, when she read a paper describing the use of one of her subject enzymes as a suicide gene. “I immediately took my mutants and said, ‘We can do better with these.’ It totally shifted in two minutes what I was doing already. It was just a little bit of an extension, and it made a big difference.”