Design by Evolution: Bringing New Chemistry to Life

In less than a month, the Nobel Prizes will be announced again. Of course, like every year, we are very curious to see who will receive this highest award. Since the first Nobel Prize in Chemistry was awarded, it has been awarded 111 times. A total of 183 researchers have received the Nobel Prize in Chemistry, including 178 men (97.3 %) and five women (2.7 %). In the future, we will presumably see more female Nobel Prize winners, hopefully as soon as this year.

In 2018, the Nobel Prize in Chemistry was divided, with one half being awarded to Frances H. Arnold, Professor at the California Institute of Technology (Caltech), Pasadena, USA, “for the directed evolution of enzymes“. I had the opportunity to meet with Arnold via Zoom and attended her online presentation and Q&A session with young researchers from all around the world at the Online Science Days 2020 organized by Lindau Nobel Laureate Meetings. On a screen, you can see the face of the speaker much better than in a large lecture hall. So perhaps even better than in a lecture hall, I could see how much Frances Arnold enjoyed talking to the next generation of scientists and perceive her enthusiasm for her research.

Frances Arnold describes herself as a “problem solver” and an engineer who is inspired by the “marvelous world of biology”. She thinks “biology is a very good place to look for some of the solutions to important problems of the 21st century”.

When asked to explain what directed evolution of enzymes is, Arnold says she breeds enzymes in a similar way to how we breed plants or animals to be more efficient or look nicer. In nature, for example, no enzymes exist for catalyzing the formation of chemical bonds not found in nature. She pioneered methods of directed evolution to create enzymes that, for example, catalyze the formation of such bonds not found in nature but that are important in chemistry. It is amazing to hear how well these engineered enzymes do their new job.

1 Directed Evolution Entering Synthetic Chemistry

1.1 Nature’s Amazing Potential

Back in the early 1990’s, Arnold tried to change the enzyme subtilisin in such a way that it would catalyze chemical reactions in a mixture of water and an organic solvent [1]. She created mutations in the enzyme’s genetic code and then used bacteria to produce thousands of different variants of subtilisin. She selected the variant of subtilisin that was most effective in the organic solvent and subjected this variant again to mutations. After several “generations”, a useful enzyme for the desired purpose had been created. Nature learned, as Frances Arnold would say. Such methods, which rely on evolution instead of “rational” human design, are today routinely used to develop new catalysts.

A more recent example in Arnold’s research is a heme protein that catalyzes the formation of organosilicon compounds under physiological conditions through a carbene insertion into silicon–hydrogen bonds. By using directed evolution, she and her team enhanced the catalytic function of cytochrome c from Rhodothermus marinus, a thermophilic bacterium from hot springs in Iceland whose native function has nothing to do with catalysis [2]. Its native function is electron transfer.

When the protein is put inside a bacterium, such as Escherichia coli, that expresses this gene, it chemoselectively catalyzes the formation of a new carbon–silicon bond. It does not do this at an equally reactive nitrogen–hydrogen bond but “chooses” the one that you have trained it on.

“Over just three generations of this process, we improved the activity from the tiny bit that we found in the cytochrome c, about 40 turnovers, to more than 1,500, which was many times better than the best human-made catalyst ever reported for this reaction. That’s quite remarkable: the protein not only could do this, but it could do it and “learn” how to do it better through evolution”, Frances Arnold explained in her talk.

With such results, it is not difficult to understand why she is fascinated with enzymes and how these dynamic systems evolve in real time. “I love this idea that nature can learn, that you can teach biological systems to do new chemistry. That we could encode in DNA chemistry that human beings invented. And chemistry they have not even invented yet. We can explore a whole new space of function that no one thought was possible. I think that is fun to be in a place where no one else has been before.”

Frances Arnold is an engineer by training. Maybe that’s part of the secret of her success. “I came at the protein engineering problem from an engineering perspective, free from the rigor that biochemical scientists felt the need to use. Rather than try to minimize the complexity of proteins by developing a ‘rational’ design approach, which would require that I understand the system in detail, I used what has worked well for billions of years: evolution. I use technology to speed it up a bit.”

1.2 Skeptical Chemists

Arnold complains, however, that biocatalysis is not yet established well enough. There are many chemists that prefer to use conventional catalysis rather than enzymes. They have no idea that “biology can do human-invented chemistry and can do it even better than we can do it”. Many chemists lack any experience with enzymes. “And then they come to my lab and see how easy it is and wonder: Why didn’t I do this before?“

The field of directed enzyme evolution is still at its beginning. Research-wise, it is scratching the surface of what is possible. Education in this field is also just starting out. Synthetic chemists do not yet have enzymes in their standard toolbox and do not know enough about the possibilities of genetically encoded chemistry that goes beyond what has been seen in the biological world.

One of Arnold’s future visions is to see synthetic chemistry and biology married. “We will not be able to put all of synthetic chemistry into biology, but, likewise, we cannot do all of biology in synthetic chemistry”, in her opinion.

Nevertheless, the number of those who use directed evolution is increasing. Currently, many biocatalysis groups the world over are training synthetic organic chemists in these methodologies. Textbooks are picking up on it. The pharmaceutical industry is moving towards biocatalysis groups. This allows them to replace toxic metals, organic solvents, and lots of waste materials and come up with cleaner, cheaper processes for manufacturing pharmaceuticals. Arnold hopes “the pull from industry and the push from education will give us a future where biocatalysis will play a major role in synthesis”.

2 Directed Evolution Supports a Sustainable Future

In her work, Frances Arnold is inspired by the beauty of biology and driven by curiosity. She loves what she is doing and says she would even do it for free. Since her early studies, when she worked on solar energy, she has been concerned about the health of the planet and interested in green chemistry.

“I was motivated many years ago by the desire to do chemistry in a clean, green, and efficient way. Chemistry is so important to our daily lives: clothing, food, materials, housing, basically everything we use in our daily lives comes mostly from chemistry. We have to find a way to do it better than what we have been doing, because we will soon have ten billion people on the planet, and I happen to like the planet that we are on and all the natural world that we enjoy. So we have to be more efficient, less wasteful, remove toxic components, and biology—enzymes in particular—do that very well.”

We should let nature inspire us. Most things in nature are recycled; they are eaten. In the past, our chemistry has not been very clean. We create a lot of waste. We use toxic metals that have to be mined, which creates pollution and degrades our environment. We all know what is wrong, but to solve it we need to revolutionize the way we do chemistry. We need science, and we probably need the smartest minds to work on these problems. In Frances Arnold’s opinion, evolution can help here. It is chemistry “done in water for the price of sugar and sunlight”. “If we can learn how to harness nature’s powerful optimization and innovation processes to do chemistry in a new way, we have the hope of opening up a clean, renewable world of chemistry for the future. We have just scratched the surface of what nature can do.”

“The wonderful thing about evolution is that it works at all scales from molecules to ecosystems.”

3 The Role of Machine Learning in Directed Evolution

Directed evolution generates a lot of data: In an iterative process of model building and experimentation, data from each generation are taken to build models, they are fed back, and it is then decided how to make mutations in the subsequent generation. Machine-learning-guided directed evolution makes this process more efficient and enables the optimization of protein functions. The expense of experimentally testing a large number of protein variants can be reduced and the result is improved.

In her current research, Frances Arnold tries to jumpstart evolution. Using machine learning, for example, she is looking for better enzymes with multiple mutations in one step. “We feel that the rare beneficial combinations of mutations can be captured by looking at the data and predicting epistatic effects.” Epistatic effects are interactions between mutations. “We are currently looking at making four or five mutations simultaneously and exploring that fitness landscape with machine learning.”

4 Designing Function

Frances Arnold says she is breeding enzymes. Currently, we do not fully understand the effects of all the mutations that we can generate. There is a large gap between our ability to sequence genomes and determine protein structures and our capacity of linking these structures to their function. But structure is not function. We are not able to design function. Especially when it comes to chemistry.

The question of when this will be possible is, of course, asked again and again. Arnold used to predict that this was not going to happen in her scientific lifetime. But now she replies: “Maybe I will have to retire soon to keep my statement true. But I am not too nervous about the next five, ten years.”

5 Women, Mentors, and Role Models

Frances Arnold studied mechanical and aerospace engineering in the 1970s. She was one of very few women in her field then. Even among the Nobel Laureates in chemistry, she is unfortunately still an exception as a woman. She explains to young female scientists, “I never let that stop me because I was just as good as everybody else was.” And she continues: “I had supportive male colleagues throughout my career and ignored those who were not supportive.”

Nevertheless, sometimes it is difficult to enter a field when there are no role models. She advocates female scientists to tell their stories so that we can create these role models and exchange experiences.

Her father, a nuclear physicist, was a very important figure in her life. “He told me I could do anything I put my mind to and to strive for excellence”. She had some mentors who have supported her, among them many colleagues and friends at Caltech, where she has been for more than 30 years. She describes them as “those who were kind enough to criticize my work in a way that I could hear and those who encouraged me to take on the hardest problems.”

An exciting question is who a Nobel Prize winner admires. For Frances Arnold, these are “scientists in their 90s who still go to work and who are joyful with their science.”

Normally, she travels a lot. Due to the pandemic, most of her work is now done online, and she has more time at home. Frances Arnold likes cooking and spending time in her garden with lots of vegetables. She also enjoys her little cabin in the San Gabriel Mountains, near Caltech. It has no water and no electricity, so it sounds like a place where you can take a real break (except for when the bears decide to visit).

Frances Hamilton Arnold, born in Pittsburgh, Pennsylvania, USA, in July 1956, studied mechanical and aerospace engineering at Princeton University, USA. After receiving her B.Sc. in 1979, she worked in Brazil and at the Solar Energy Research Institute in Colorado, USA. She attained her Ph.D. in chemical engineering from the University of California, Berkeley, USA, in 1985. After a postdoctoral year in biophysical chemistry at UC Berkeley, she joined the faculty at California Institute of Technology (Caltech), Pasadena, USA. Currently, she is the Linus Pauling Professor of Chemical Engineering, Bioengineering and Biochemistry at Caltech.

Arnold’s research focuses on engineering enzymes to catalyze reactions not known in biology and developing new machine-learning-guided protein evolution methods.