03/27/1996 08:00:00

Chemical Engineers Show that Directed Evolution Can Be Useful

PASADENA—Caltech engineers have shown for the first time that an experimental technique known as directed evolution can solve real, industrial problems in pharmaceutical manufacturing.

The result, published in the April 1 issue of Nature Biotechnology, describes how the researchers used directed evolution to develop a new enzyme that is able to catalyze—increase the reaction rate of—an important step in the manufacture of an antibiotic.

"We're very excited to demonstrate that the technique works on a problem of real, industrial interest," said Frances Arnold, an associate professor of chemical engineering at Caltech. She collaborated on the research with Jeffrey C. Moore, a graduate student in chemical engineering.

The production of most chemicals, including antibiotics, requires many steps, some of which are quite slow. To speed up these steps, chemical engineers use catalysts, which can be metals, metal-based molecules, or ordinary enzymes.

In the present case, researchers were working on a step in the production of a class of antibiotics derived from cephalosporins. The pharmaceutical company Eli Lilly, while developing the manufacturing process for these antibiotics some years ago, found that zinc catalyzes this step effectively, but the zinc process is expensive, and they wanted to use something less costly.

Company scientists thought a naturally occurring enzyme might catalyze the reaction, in this case cutting molecules called p-nitrobenzyl esters into two smaller pieces, one of which forms the basis for an antibiotic called loracarbef. After screening many enzymes they finally found one that catalyzed the reaction, but it was only weakly effective, so they suspended research on the project.

In 1994 Eli Lilly gave the enzyme to Arnold, to see whether her lab might improve it using directed evolution. Directed evolution is a process in which the gene that produces a natural enzyme is mutated so that it produces variants of the enzyme. Several generations of mutation and screening for enhanced activity can significantly improve the enzyme.

Enzymes are made of strings of hundreds of amino acids, and the mutated genes typically create enzymes with several of the many amino acids replaced by different ones. Substituting amino acids can change the shape and function of the enzyme. Some variants are less effective, some are more so. Researchers take the new, variant enzymes and test them for the desired activity. The genes that produce the best, most active enzymes are saved and they are either further mutated or recombined, and the new variants are screened again for their effectiveness. At each step, scientists save the genes that produce the most active enzymes to become the "parents" for the subsequent generation.

To create an industrially useful catalyst, Moore and Arnold needed to increase the enzyme's activity by at least 10 times. They far surpassed their goal. After five generations of mutations, recombinations, and screenings, they found enzymes that are as much as 30 times more active than the one they started with.

As it turns out, the new enzymes won't be used commercially yet, because the pharmaceutical company had already received approval from the Food and Drug Administration to manufacture the antibiotic using the zinc catalyst. And though the zinc process is expensive, it would cost the company even more to go through the FDA approval process again.

But that's of little concern to Arnold. "Of course it would be nice if we had started work on this novel enzyme earlier, so that it could have been used commercially," Arnold said. "But it may be used for other products. And, more important, we've shown the capability of this technique to solve real problems. I'm sure it will find many useful applications in the future."

Contact: Jay Aller (818) 395-3631 aller@caltech.edu

Written by John Avery