How does directed evolution improve enzymes?
You start with a gene that codes for an enzyme with some baseline activity—maybe it's mediocre at breaking down cellulose or synthesizing a drug intermediate. You then create a library of mutant genes, each with random changes. The trick is to generate enough diversity: typically millions of variants. Then you apply a selection pressure—for example, growing bacteria that need the enzyme to survive, or using a high-throughput screen that detects product formation. The variants that perform best under that pressure are isolated, their genes are amplified, and you repeat the cycle. Each round pushes the enzyme further up the fitness landscape. I've evolved enzymes to withstand temperatures above 70°C, to work in organic solvents, and to catalyze reactions that don't exist in nature. The beauty is that you don't need to understand the mechanism in advance—you just need a good assay. Nature has had a 3.8-billion-year head start, but we can accelerate that in the lab.
Ask Frances Arnold the follow-up →