Peng Liu is unraveling new reactions — and new ways of working together

Publishing in Nature is a crowning achievement for any researcher, marking the kind of work that can make a career. Peng Liu and his students managed it twice earlier this year — in the same month.

The Kenneth P. Dietrich School of Arts and Sciences professor is on a hot streak that shows no sign of stopping, and those two papers — though just a small sample of the many recent high-profile studies his team has published — serve as a guidepost for how he does it. One details a totally new chemical reaction that could find use in new therapeutics, while the other rewrites a textbook reaction that’s taught in every introductory organic chemistry class. 

“The research has been very exciting,” said Liu. “We now have a collection of computational chemistry methods that allows us to tackle problems in completely different types of systems.”

Unraveling new chemistry

The first study, published in Nature earlier this year, described a brand-new chemical reaction that, on paper, has all the hallmarks of being challenging to control. 

“There’s a famous principle in organic chemistry, even if it’s not always true: If something is more reactive, it’s usually not very selective,” Liu said. The reaction the team described makes use of famously reactive “free radicals” as an intermediate step and acts on two different parts of a molecule that can rotate freely, making it even harder for a chemist to ensure that the final product is in just the right shape.

Using an engineered light-sensitive enzyme to shepherd the required components into just the right formation, the team could ensure that they got the end products they were looking for while avoiding the use of toxic materials or extra steps. The result is a new way to make certain kinds of amino acids that don’t exist in nature, potentially opening new avenues to create protein-based therapeutics. But the method also created a challenge for Liu’s team. 

Liu is a computational chemist who teases apart and provides insight on the discoveries of colleagues who do their chemistry in a lab. So when his collaborators at the University of California, Santa Barbara brought him the results of their lab work, it was the Pitt team’s job to figure out what was going on under the hood. The reaction was both new and complex, requiring an innovative technical approach to understand. 

“We pushed to combine two separate types of computational techniques,” Liu said. Molecular modeling enabled them to figure out what’s happening at the broad scale, made necessary by the use of enzymes, which can contain thousands of atoms. The more computationally intensive quantum modeling let them zero in on how specific chemical bonds form and break with more precision. 

By marrying disparate approaches and finding new ways to peer into the inner workings of chemical reactions, Liu makes it easier for other researchers to expand on the work — and maybe find a way to some new chemistry of their own.

“Combining these methods is going to be very helpful to study complex catalytic reactions,” Liu said. “The question is always: How can we accelerate the next reaction, and how do we ensure the desired product is formed?”

The importance of strong bonds 

Another theme of Liu’s research is a spirit of collaboration and communication. In the second Nature paper published earlier this year, Liu relied on his students Christian Knox, Mithun Madhusudhanan and Thomas Tugwell, who collaborated closely with an experimental team under his guidance. Working this time with researchers from the University of Chicago, Liu and his students showed a new way to, with unusual precision, build long chains of a class of useful molecules that have found use as chemotherapy drugs and other therapeutics.

Also unusual was the reaction mechanism that Liu and his group proposed, cracking open a foundational piece of organic chemistry called the SN2 reaction.

“This is a reaction we teach in our sophomore undergraduate classes, a reaction over a thousand Pitt students learn every year — and this paper changes all that,” Liu said. “It’s a fairly bold claim that this is the actual mechanism.”

Extraordinary claims require extraordinary evidence, so Knox, Madhusudhanan and Tugwell worked with the experimental team to show the mechanism’s plausibility with two completely different methods of modeling. It was a feat made possible only because of the two labs’ long working relationship. 

“We interact very closely with our experimental collaborators,” Liu said. “I think this is really the key to our success and how we can apply our methods to so many different areas and study so many different reactions.”

As the team builds more sophisticated modeling techniques and gains access to more computational power — with support from Pitt’s Center for Research Computing — Liu sees the potential for these collaborations to become even closer. Rather than experimentalists handing Liu a single chemical reaction to figure out the mechanism, why not apply the same computer modeling methods from the beginning to guide the process when lab-based teams are screening hundreds of different reactions to avoid dead ends?

“We’ve always been trying to facilitate all stages of experimental development,” he said. “It’s a completely different type of challenge to predict new chemical reactions, but that’s an important direction of our research.”

[Read about yet another high-impact paper the team co-authored last month, this time in Science.]

Wielding both a Swiss army knife of modeling methods and a deep rolodex of collaborators, Liu’s lab is well-placed to make this new interplay between theory and practice a reality.

“Hopefully, there’ll be a lot more exciting systems that we can push our computational techniques to get more insights and look at things that are even more complex,” Liu said.

Photography by Tom Altany