Take a piece of living matter and make it work better.
That’s what Frances Arnold is trying to do in a few weeks in her laboratory–what nature usually needs many years to accomplish.
A 41-year-old biochemist at the California Institute of Technology in Pasadena, Arnold is re-enacting the course of evolution at warp speed in order to develop new medicines and materials–even a better laundry detergent–faster and cheaper than is now possible.
She’s a leader in the budding science of artificial or “directed” evolution, invented less than four years ago. The idea is to imitate the natural processes described by biologist Charles Darwin in his pioneering “Origin of the Species” 140 years ago. First comes mutation to create variations, then selection of the best variants, then breeding or sexual recombination to produce a new, improved generation.
A swifter pace of evolution is of particular interest in the pharmaceutical industry, where the first company to market a new drug enjoys an enormous competitive advantage. It typically takes 10 to 12 years and costs $100 million to $300 million to develop a drug, according to Thomas Caskey, senior vice president for research at Merck Research Labs in West Point, Pa.
Besides the economics, rapid evolution helps the drug companies keep up with their enemies: bacteria and viruses that develop resistance to antibiotics by rapidly mutating themselves.
Artificial evolution is sometimes called “molecular breeding”–a scientific euphemism for sex in a test tube. Just as selective breeding can improve the qualities of dogs, flowers, cows or rice, biologists are mating microscopic molecules to generate better products.
“We start with something functional and vary that by random variation, as nature does,” Arnold explained in a recent lecture to a science conference in Philadelphia. “We identify the beneficial effects of small changes and accumulate them. We’re mimicking natural evolution over hundreds of millions of years.”
In her laboratory, Arnold works with enzymes, proteins that change one kind of chemical molecule into another. Proteins are chains of molecules that form the building blocks of all living matter. The instructions to make a protein are carried in the DNA coiled up in the genes contained in every cell. A tiny change in a gene’s DNA is enough to modify the protein or enzyme–sometimes for better but usually for worse.
“Enzymes are masters of chemistry,” Arnold said. “They evolved over billions of years to perform specific biological functions. They make complex materials with virtually no waste.”
Trouble is, enzymes are so narrowly specialized that it’s hard to get them to do something different. Sometimes they shut down too soon or refuse to work under difficult conditions, such as heat or acidity, common in industrial settings.
Enzymes needed to produce cephalosporins, a widely used family of antibiotics, tend to break down under high temperatures. As a result, Eli Lilly & Co., the big drug manufacturer based in Indianapolis, asked Arnold’s lab to help it develop a sturdier enzyme.
Imitating natural evolution, Arnold plucked enzyme genes from a cell, broke them up and zapped them with chemicals to alter their DNA. She transplanted the variant enzymes into a set of bacteria, then she “roasted” the bacteria, picked out the ones that withstood heat best and repeated the process. After five generations, she said she had bred a strain of enzymes that were stable at temperatures up to 140 degrees Fahrenheit and performed 100 times better than the original.
“We were very impressed with her science,” said Andrew Russell, director of bioprocess research and development at Eli Lilly. “With this technology, she is able to significantly improve some properties of enzymes that will be useful to Lilly in making a drug for sale.”
Other businesses besides drug companies are also interested. Procter and Gamble of Cincinnati is drawing on Arnold’s work to develop a laundry detergent that will dissolve dirt at higher temperatures. The Dow Chemical Co. of Midland, Mich., is seeking new ways to make useful products out of oil-based substances.
According to Arnold, it is far more effective to let artificial evolution take its course than to try to consciously design a new protein.
“These are horrendously complex molecules,” she said. “We don’t dictate what the solution will be. If we use the design process that nature teaches us, we’re going to get answers that are much more interesting than our puny brains can do.”
