Rather than seeking to design new proteins rationally, piece by carefully calculated piece — as many protein chemists have tried and mostly failed to do — the Arnold approach lets basic evolutionary algorithms do the work of protein composition and protein upgrades.
You start with a protein that already has some features you’re interested in, such as stability in high heat or a knack for clipping apart fats. Using a standard lab trick such as polymerase chain reaction, you randomly mutate the gene that encodes the protein.
Then you look for slight improvements in the resulting protein — a quickened pace of activity, say, or a vague inclination to carry out a task it wasn’t performing before, or a willingness to operate under conditions it deplored in the past.
You mutate the improved version again and screen the output for even better performance. Repeat as needed. You do your experiments with the help of a bacterial workhorse such as E. coli, or with an exotic microbe isolated from hot springs in Iceland where temperatures can exceed 175 degrees F.
You consciously treat proteins and their carrier microbes exactly as people unconsciously treat disease microbes when blasting them willy-nilly with antibiotics: You encourage the microbes to rise to the challenge, adapt, survive.
Through directed evolution, Dr. Arnold’s lab has generated microbes that do what organisms in nature have never been known to do. Some of them, for instance, stitch together carbon, the element that defines life, and silicon, the stuff of sand, glass and computer chips but heretofore not of life (unless you are a Horta, the rock-shaped beings who famously mind-melded with Mr. Spock on “Star Trek”).
“We showed for the first time that living organisms can use their own machinery to bring carbon and silicon together to form a bond,” said Jennifer Kan, a postdoctoral scholar in Dr. Arnold’s lab who performed the experiments. “We didn’t even have to nag the protein too hard to get it to do it.”
Dr. Arnold has another favorite mantra: “Nature doesn’t care about your calculations.” Analyzing the evolved mutations that proved most effective at tweaking a protein’s performance, the Arnold team found the changes in all sorts of unpredictable places.
Natalie Angier became a columnist for Science Times in January 2007. She joined The Times in 1990, covering genetics, evolutionary biology, medicine and other subjects, and was awarded the 1991 Pulitzer Prize in Beat Reporting.
This content was originally published here.