The evolution of an enzyme engineer who changed chemistry

In a March 2026 story, Frances Arnold discusses her technique of “directed evolution” that creates enzymes with unusual abilities.

• 0:00Speaker 00

  WNYC is supported by Bayer. Science is about asking questions, testing, trying again. It's not easy, but it's relentless. It evolves, and that's why science delivers. Learn more at sciencedelivers.com.

• 0:17Speaker 01

  Hey, it's Flora, and you're listening to Science Friday. Today, we're bringing back a conversation with an absolute all-star. And it's a conversation I've thought about a lot since we recorded it. So I hope you enjoy it. In nature, enzymes are the catalysts that make much of biology work. They jumpstart chemical reactions that either wouldn't happen or would happen super slowly. They do a lot of basic biological jobs, breaking down food, building other molecules, extracting energy. But what if we could harness evolution to engineer designer enzymes that do other jobs, specific jobs that benefit us?

• 0:58Speaker 01

  Putting that idea into practice changed the game for chemistry and earned Dr. Frances Arnold the Nobel Prize in 2018. She called it directed evolution. And this is really big. Today, thousands of labs use these methods to coax enzymes into doing things no one had ever thought of. Dr. Frances Arnold is a professor at Caltech, and we'll hear about where she sees this approach going in the future and her own personal evolution, the evolution that brought her into science. Welcome to Science Friday, Dr. Arnold.

• 1:32Speaker 02

  Oh, thank you so much, Flora. I'm thrilled to be here.

• 1:35Speaker 01

  OK, in our intro, we described enzymes as catalysts that make biology work. I know they're near and dear to you. Do you have a more poetic way of thinking about them?

• 1:45Speaker 02

  Poetic? Well, I think about them as the transformation agents of the whole natural world. It's enzymes that take carbon dioxide and sunlight and, and simple starting materials and build trees and human beings and flowers and birds. They're amazing. They're the best chemists on the planet.

• 2:08Speaker 01

  When we talk about directed evolution, should I be thinking about this like enzyme breeding, but like instead of a chihuahua, you're making, you know, a tiny adorable protein?

• 2:19Speaker 02

  Well, I think of it that way. It is a way to explain that we have been modifying the biological world at the level of DNA for ages, making strange things like chihuahuas that wouldn't be found naturally. But for some reason, human beings find them charming or useful. So why not use that same very powerful process of, of directing the evolution, which is what breeding is, of enzymes to make what we want?

• 2:52Speaker 01

  I want to get into this a little bit. I mean, let's say you wanted to create an enzyme that's like a stain remover. Can you create conditions that make a stain-removing enzyme more likely? Or is it more like trial and error, and you find one that gets a little bit better at stain-removing, and then you breed that one? How does it work?

• 3:13Speaker 02

  Well, you're right, Flora. It's a process of incremental change. You don't get a whole new enzyme at once, although that may be coming. But what we do is we start with something that has a little bit of the property that you're interested in. Let's take your example of stain remover. If you start with an enzyme that has a little bit of capability to remove stains, but doesn't work in the presence of bleach and doesn't like your laundry machine, for example, then we can breed in or evolve in those desirable properties one step at a time.

• 3:51Speaker 01

  And to do that, you take the, the gene that makes that enzyme and you put it in a living organism? Is that how that works?

• 4:00Speaker 02

  Well, for example, bacteria are wonderful translators of genetic material, of DNA, into enzymes. They can read DNA and spit out the enzymes wonderfully. So we use that capability, like, you know, a bag of reagents, basically, of bacteria to make all the mutant enzymes that we'll program in. And then we search through those bacteria to find the ones that harbor the great stain-removing enzyme.

• 4:35Speaker 01

  And how do you sort of determine the winner? Like, are you putting these, these enzymes, like, in a test tube with a piece of fabric that has a stain on it and seeing which does the best? Like, what does it look like?

• 4:49Speaker 02

  Well, I don't do that. In fact, I'll tell you a funny story. Way long ago, back in the early 1990s, I had a, a grant from Procter & Gamble, who wanted me to make an enzyme for all-temperature cheer. Their laundry detergent, and it's supposed to work at all.

• 5:06Speaker 01

  This is all too real.

• 5:07Speaker 02

  It's really real.

• 5:08Speaker 01

  My example here.

• 5:09Speaker 02

  But it's a wonderful example, because I didn't have any way to test whether it would take stains off of clothes. So I said, sure, I'll breed you an enzyme that works at all temperatures, but I'll bet it'll forget how to take stains off of clothes. Because if you breed one property, you can very well lose other things that you're interested in. So you need to be able to measure all of the properties that you're interested in. It's called the first law of directed evolution, which is you get what you screen for. And that may sound like a, you know, a silly statement, but it's actually a very profound statement, because if you don't screen or search for the thing that you really care

• 5:51Speaker 02

  about, you'll get what you searched for, but it may not be what you want. We have these great T-shirts in my lab that says, "You can't always get what you want" when we're searching the non-biological world.

• 6:04Speaker 01

  I think we can all relate to that.

• 6:06Speaker 02

  Yes.

• 6:07Speaker 01

  I mean, if you get what you screen for, do you also ever benefit from serendipity? Like, do you ever find, OK, this turns out not to be a great stain remover, but wow, it's wonderful for fermenting yogurt?

• 6:21Speaker 02

  So, yes, because that's exactly what evolution does. But if you, if you expand the space of biology and you go out into new chemical frontiers, what happens is that those new molecules pick up properties that may be interesting for something else. And that's how innovation happens in biology, how something wholly new is created. Say, an enzyme that catalyzes a whole new chemical reaction may become possible when you go out into these new spaces. But you have to know what you're looking for, and sometimes all those new properties

• 7:04Speaker 02

  would be lost because you have no way of knowing that they're there.

• 7:07Speaker 01

  So can you screen for multiple properties at once?

• 7:10Speaker 02

  Yes, if you know what you're looking for. It's not going to raise its hand and say, oh, I can, I can build a new carbon-silicon bond. I may not take off stains, but I can do this. No, it's not going to tell you that. You have to ask it a million questions.

• 7:28Speaker 01

  Well, what if you don't know the right questions to ask? What if you can't think up the right question?

• 7:33Speaker 02

  Wow. I mean, we're searching on almost infinite space, and you can't search infinite questions against an infinite space. Humans can't do that. But you know biology does it. Novelty happens because all of this incredible diversity of the biological world is constantly changing and constantly looking for new opportunities for survival, right? Survival with a fetish. So a new solution to, let's say, eating up an environmental pollutant can just pop up anywhere in the world. That's so incredible, because it gives, if you, say, for example, you break down an

• 8:15Speaker 02

  environmental pollutant and it gives you a new, rich nitrogen source, you have a selective advantage over your neighbors, and you grow faster, and then you can occupy the whole world to eat up that pollutant. I don't have any way to do that when I'm breeding in the laboratory.

• 8:32Speaker 01

  So are you starting from scratch each time? Or do you have a library of enzymes that you start from? And are you adding to it with, you know, enzymes you find in the natural world?

• 8:43Speaker 02

  All of that, yes. So we, so we have a great big refrigerator of things that we've made that are now no longer what you would find in the natural world. We have an even bigger database of enzymes that all of science has deposited, what we understand, and we have an almost infinite size of space of all possible enzymes that no one has explored.

• 9:12Speaker 01

  Can you evolve an enzyme to do something sort of completely outside of the bounds of what life normally does? And like, are there problems that are easier to solve with enzymes than other problems?

• 9:24Speaker 02

  Definitely, yes. So if you, what does evolution do? Evolution takes what is there in one form or another and makes it better and confers maybe slightly new properties. So if you know how to look at what is there, let's say, in the biological world, or what we could even design, evolution lets you mold that into something wonderful, something that actually, you know, for example, would make money or would solve a problem. So the problem for the researcher like me is to understand from what's there, what new properties could come about. And that's called chemical intuition.

• 10:10Speaker 01

  And is it intuition? I mean, is it a gut feeling?

• 10:13Speaker 02

  No, it's a lot more than than a gut feeling. It's knowing what chemistry might be possible from the machinery that already exists.

• 10:24Speaker 02

  So we know the structures of hundreds of thousands of enzymes now. And we can predict the structures of millions and millions more thanks to these new AI tools. And if a chemist looks at these structures and recognizes some machinery that might look like some chemistry that was invented by human beings, then maybe we can find this ability in those structures that evolved to do something completely different. And that means we can find novelty by being the breeder, by recognizing what the biological world might be capable of, based on our experience.

• 11:09Speaker 02

  What's even more exciting is that AI tools for design might even give us access to virtually any chemical transformation that an enzyme could do. And that is so exciting because if we could actually design a starting point for desired chemistry, not have to already find it in the biological world, then we can essentially press a button and genetically encode any chemical transformation. Wouldn't that be incredible?

• 11:44Speaker 01

  I mean, it sounds truly incredible, like beyond belief.

• 11:48Speaker 02

  But it's not beyond belief. All the steps towards that are being taken today. AI design of enzyme, obviously, the evolutionary optimization, novel measurement techniques to let us sort through all of these new genetic sequences. I think in the next five to 10 years, we will be able to genetically encode almost all chemical transformations that you could perform in an enzymatic system.

• 12:19Speaker 01

  In the next five to 10 years, you could type in, like vibe coding. I want this chemical transformation.

• 12:26Speaker 02

  Yep.

• 12:26Speaker 01

  Make me an enzyme.

• 12:27Speaker 02

  Yep, I think so.

• 12:29Speaker 01

  You heard it here first, everybody.

• 12:33Speaker 01

  We need to take a break, but there's so much more to talk about. When we come back, turning from, you know, what you achieved, Frances Arnold, to why the evolution of a scientist. Stick around.

• 12:55Speaker 01

  Hey, everybody. If you like Science Friday, I want to recommend another podcast I know you'll love to, Science Versus. From peptides to protein-maxing, Science Versus takes on fads and trends and the opinionated mob, I know you know what I mean, to find out what is fact, what is not, and what's something in between. And they're the only podcast with better pun game than us. So check them out, Science Versus, wherever you get your podcasts.

• 13:28Speaker 01

  Francis, can we get personal for a second?

• 13:30Speaker 02

  Sure.

• 13:33Speaker 01

  You said that in a way like, no, thank you.

• 13:37Speaker 02

  No, no, that's fine. Everything that's ever been said to me is now out in ChatGPT.

• 13:46Speaker 02

  Why not add a few more?

• 13:48Speaker 01

  I want to know what you were like as a kid. Give me a sense of young Francis.

• 13:53Speaker 02

  Oh, young Francis was a real pain in the you-know-what. I was bored in school and so was off doing my own thing most of the time. I was very curious about the world and so didn't pay attention to the very boring classes that we had, starting in elementary school. So you know what they did with a little girl in the 1960s who was bored in school? They sent me to typing class. So at age 10, I learned how to type, sitting on phone books. Learned how to type, and then I took mechanical drawing, geometry, and a number of other classes in the high school. And so that by the time I went to middle school,

• 14:38Speaker 02

  I was completely bored and I started cutting classes. I didn't go to school anymore. This started in eighth grade. I got kicked out of the private school that my mother had worked so hard to get me into because they would cater to bright girls. But I got kicked out because I would go off and do other things instead of going to school. And I just said, OK, well, high school is not very interesting, and started leaving school again.

• 15:07Speaker 01

  How did that go over with your family?

• 15:09Speaker 02

  Well, they didn't like it. They told me, shape up or ship out. So at age 15, I went out and got my own apartment. And I worked in various horrible jobs. I had so many bad jobs that it made me want to go to college.

• 15:25Speaker 01

  Tell me about your terrible jobs.

• 15:29Speaker 02

  Oh, you know, I've had the jobs that, if they're lucky, teenagers will continue to get, working in pizza parlors and sales department store. And then by the time I could lie about my age and pass for 21, I talked myself into a good job as a cocktail waitress at a local jazz club. And that was really fun, except for it kept me up all night and therefore I didn't go to school. And then finally, I got a job as a taxi driver, which I really enjoyed. I learned how to get around Pittsburgh, Pennsylvania, which is a very complicated city, in one of those fat, old, yellow cabs with no GPS. I want you to think about that.

• 16:14Speaker 02

  There was no GPS then.

• 16:16Speaker 01

  Isn't there a part of the brain that expands for taxi drivers because they keep these mental maps in their head?

• 16:22Speaker 02

  I think so. I think it was a wonderful experience and I loved learning all about Pittsburgh.

• 16:28Speaker 01

  I read that you had to hide your hair under your hat because people didn't want a woman driver.

• 16:33Speaker 02

  Well, you can imagine. This was, what, 1974, I think? People were not used to seeing women behind the wheel of a big, fat, yellow cab.

• 16:46Speaker 01

  I can imagine, yeah.

• 16:47Speaker 02

  So once they got in, then they got in the car, then it's too late to leave.

• 16:54Speaker 01

  I mean, you talk about your rebellious streak. And I think, you know, many really successful scientists I've spoken with have that characteristic. Do you think that's necessary for making a paradigm-shifting discovery?

• 17:11Speaker 02

  I think what's necessary is the willingness to go beyond what all the other monkeys are doing. This is what I tell my children. Why are you doing what all the other monkeys are doing? You have to be willing to do something that is outside the bounds. Otherwise, you'll just be adding incrementally, at best, to what's already there. So it takes a sense of adventure, maybe.

• 17:39Speaker 01

  And a little bit of courage, I think.

• 17:41Speaker 02

  Oh, definitely courage.

• 17:43Speaker 01

  Where do you get that from?

• 17:44Speaker 02

  I never cared what people thought. So I've been gifted with the ability to let criticism go in one ear and depart from the other one without getting too worked up about it. And is that courage? No, it's just I followed what I wanted to do rather than what other people thought I should do.

• 18:05Speaker 01

  Is there a moment from your childhood that was predictive of your career in science? I mean, you've done so many things.

• 18:13Speaker 02

  Oh, I don't think of that, no. There's no moment that's predictive. And whose career is predicted? I don't. I find it very hard to predict even the next five years, even though I just did that with the prediction that we'll be able to genetically encode chemistry. But I can predict that because I can make it happen.

• 18:36Speaker 01

  You know, I want to talk a little. I know you're interested in the humanities, too. I have to ask, how many languages do you speak?

• 18:43Speaker 02

  I can understand French very well. I lived in Italy for a couple of years, so I speak pretty good Italian. I speak Spanish. I lived in Madrid. I used to speak Portuguese because I lived in Brazil, but that has pretty much fallen by the wayside.

• 18:59Speaker 01

  That's a lot. You mentioned also Yiddish. And you seem to have a flair, like, for languages. I mean, do you see that interest as separate from your scientific interests or do they connect in some way?

• 19:11Speaker 02

  I think it's a connection. I'm just curious about words. I'm as curious about words as I am about DNA. And the evolution of languages is completely fascinating, just like the evolution of proteins. You can build family trees of languages and have always been thrilled by how words express thoughts in different languages.

• 19:37Speaker 01

  You know, reading your Nobel biography it almost reads like fiction, just the number of experiences you've had, traveling across Europe, sleeping on farms, off-road motorcycling, taking care of horses. Do you feel like your life could have gone in many different directions?

• 19:53Speaker 02

  Oh, of course. I think all lives can go in different directions. And just like evolution, there's a contingency there. This opportunity came in front of me at this time, and I chose this path, whereas I could have had a different opportunity at a different time. And the trick, I think, is to choose a path, is not be paralyzed by opportunities. Choose something and go for it. And you can always change directions if you don't like it. So I tried many things before I found what I loved. At age 30, I finally found what I like. But I did a lot of things before then.

• 20:38Speaker 01

  Let's fast forward to your career in science. So you work at the intersection of biology and chemistry and engineering. Did you have to carve that niche out?

• 20:48Speaker 02

  Yes, I did. I started as an assistant professor of chemical engineering at a time when chemical engineers were all working for the oil companies. I thought I would work on alternative energy because I had come from the solar energy background. But I ended up at Berkeley getting my PhD at the beginning of the DNA revolution. Imagine how exciting that was. So here I am, an engineer, but I could see that there was this future of rewriting the code of life. And I said, I want to be an engineer of the biological world at a time when that first became possible.

• 21:32Speaker 01

  Right, because it didn't exist before.

• 21:33Speaker 02

  It didn't exist. And my timing was perfect.

• 21:38Speaker 01

  It's timing, but it's also you saw the opportunity.

• 21:42Speaker 02

  I did. I did. But I was lucky to be in the middle of it.

• 21:47Speaker 01

  What was the biggest failure in your career?

• 21:50Speaker 02

  Oh, I've had so many failures. The biggest failure is staying too long at something. Or I don't dwell on failures. I just move quickly to other opportunities. So I don't actually think about big failures very much.

• 22:10Speaker 01

  Well, I mean, I wonder if that's sort of key to success.

• 22:14Speaker 02

  Well, of course, that could be. But sometimes those who dwell on failure learn something really, really hard and really important. So there's probably some sweet spot, right? To work hard at a problem and achieve it or give up and move on to the next thing. And sometimes you have to do one, sometimes you have to do the other.

• 22:40Speaker 01

  You know, you were often out front as one of the few women working in your institution or your department. You are one of eight women to win the Nobel Prize in Chemistry. Going back to your taxi days, did you ever feel like you needed to hide your hair under your hat in your scientific career?

• 23:00Speaker 02

  No, never.

• 23:03Speaker 02

  I took advantage of it. Because scientists, just like everybody else, want people to listen to them. Every scientist is working hard to solve a problem, and we love talking about it. Even the people who are shy love talking about it. And if you go out, say, to a conference and you give a talk and nobody comes, that's kind of sad, isn't it? Well, I was an oddity. And so people, when I walked in the room, they'd sit up and say, oh, my goodness, who's this? And they would listen to me. And I just made sure that the first words that came out of my mouth caught their attention,

• 23:47Speaker 02

  and they would stay to listen more. So I took advantage of this oddity that I was to get in front of people with my ideas.

• 23:59Speaker 01

  That's really interesting to me. I mean, partly just also in the way you think about the power of language, right? Like, I think scientists are often very careful in the way that they talk. Not a lot of metaphors. Yes, a lot of caveats. What is your philosophy on that?

• 24:18Speaker 02

  My philosophy is that scientists need to tell a story. It doesn't have to be embellished. It doesn't have to be hyperbolic. But it should be a story that captures the attention of the audience. Otherwise, you put people to sleep. We're very good at that, as you probably well know. And I decided early on I didn't want to put people to sleep and that I wanted to share the beauty of the enzyme world and my excitement about evolution. And honestly, it is truly exciting and wonderful to hear the stories.

• 24:59Speaker 01

  Well, let's talk about the next chapter of your work and of this field. You know, I heard from a biologist recently that the 21st century, he said, it's going to be biology's moment. Like, this is biology's moment to sort of join the grown-up table of science, where it's typically been physics and mathematics and chemistry. And I know that you're working at the intersection. But do you agree with that sentiment?

• 25:26Speaker 02

  Well, I think it'll be science's moment. As big AI turns its attention to science, we will have incredibly powerful tools to address our lack of understanding of the physical and the biological world. I think biology is particularly exciting because it's so complex. And the new tools that are coming up will help us navigate this complexity and get better understanding, at least, I hope, get better understanding of the complexities of life.

• 26:06Speaker 01

  What are some of the problems that you're working on right now?

• 26:10Speaker 02

  Well, the one that I mentioned at the beginning, which is how do you genetically encode any chemistry? So half of my research team at Caltech is expert in AI and machine learning. I'm not.

• 26:27Speaker 02

  I kind of ride along the top. And we are integrating these experimental methods of evolution in the laboratory with AI tools so that we can explore this universe of possibilities that biology provides.

• 26:44Speaker 01

  It's interesting because you're not focused on one problem. It's like you're focused on all the problems, finding a solution for all the problems.

• 26:52Speaker 02

  Well, we do it in the context of individual problems. So we'll take on a challenge of making, for example, a tuberculosis drug using enzymes so that it can be sold cheaply in developing countries. That's a really hard enzyme design problem. And we took money from the Gates Foundation to build enzymes that do new-to-nature chemistry to do that. And it worked really quite well. Or we have a project to degrade forever chemicals using enzymes for which there's no known chemistry, but we can create that in the laboratory. Those are examples that allow us to test our methods

• 27:36Speaker 02

  and see if what we say we can do, we can actually do.

• 27:41Speaker 01

  Do you think Forever Chemicals will be kind of like a killer app for designer enzymes?

• 27:46Speaker 02

  I hope so.

• 27:50Speaker 01

  How close are we?

• 27:53Speaker 02

  It's a complicated issue. PFAS. Let me just say that enzymes are one of the many ways one could consider to attacking forever chemicals. And one of those ways, of course, is not to make them in the first place. So what I'm hoping is that enzymes will provide a clear-cut, hugely cost-effective method for degrading plastics or forever chemicals, and that they will win the race. But I hope anyone wins the race.

• 28:27Speaker 01

  If you were starting fresh today, would you be working in the same field? Or what problem would you tackle?

• 28:35Speaker 02

  I would work exactly on what I'm working. It has given me endless joy, fascination, fulfillment, interest. I just love it.

• 28:47Speaker 01

  You've found your evolutionary niche.

• 28:49Speaker 02

  I did. And young people find it exciting. I've been doing this for 40 years, and the excitement just continues to grow. It's not a field that's closing off. Many fields that people were working on 40 years ago are kind of, you know, closing off. We've answered those questions, and it's time to move on to something else. But rebuilding or building a new biological world, that's just something that's exploding in opportunities. And I'm thrilled to be at the beginning of it.

• 29:24Speaker 01

  I'm going to ask you one last question, and it's like a big, thinky, philosophical one. What's the purpose of science?

• 29:33Speaker 02

  The purpose of science?

• 29:37Speaker 02

  As a technologist and an engineer, I translate science to solving problems for the planet and human beings. But the purpose of science is to give us a better understanding of where we sit in the universe and what is our role and our origins and our future in this universe.

• 30:01Speaker 01

  Dr. Frances Arnold is the Linus Pauling Professor of Chemical Engineering, Bioengineering, and Biochemistry at the California Institute of Technology and the winner of the 2018 Nobel Prize in Chemistry. Frances, thank you so much for talking with me today.

• 30:15Speaker 02

  Thank you for having me.

• 30:17Speaker 01

  It was really a pleasure. If you want to help this podcast evolve a bit, why not give us some direction? Call us at 877-4-CY-FRY and coax us to solve some new problems. That's 877-4-CY-FRY. Call us with your questions that you can't find answers to. This episode was produced by Charles Burquist. Thank you for listening. I'm Flora Lixman. Catch you tomorrow.