The discovery and design of new chemicals have traditionally been time-intensive and labor-intensive processes, but Sijia Dong, an assistant professor in chemistry and chemical biology, seeks to change that by applying advanced simulations, AI, and quantum computing. With additional ties in physics and chemical engineering, Dong is pioneering methods that use these technologies to accelerate chemical discovery and enable innovative pathways for developing new materials, enzymes, and reactions.
Chemical reactions are governed by the behaviors of subatomic particles—electrons and protons—within molecules. Dong’s research team uses this foundational principle to investigate enzymes, proteins that facilitate chemical transformations. They focus specifically on photochemistry, which examines how chemicals react to photons, or light. In recent research published in Chem, Dong and her colleagues leverage photochemistry to drive enzymatic synthesis through a phenomenon known as charge transfer. By doing so, they aim to design light-driven “photoenzymes” that can enable energy-efficient, large-scale chemical synthesis, which holds promise for various industries, including pharmaceuticals.
This charge transfer process occurs in a molecule’s “electronic excited states,” where electrons occupy higher energy levels. By designing enzymes that can harness these excited states, Dong’s team has opened a novel avenue in enzyme engineering that is rarely explored. Through computational modeling, they discovered that by altering the protein structure, they can control the electronic characteristics of charge transfer complexes, directly influencing how these reactions proceed.
For pharmaceutical companies and other sectors, light-driven chemistry presents an appealing alternative for drug synthesis, as it could reduce energy costs and streamline production. Dong’s ongoing work centers on building a computational framework that enables the prediction, design, and mechanistic understanding of photoenzymes, which could transform catalyst design and significantly enhance the efficiency of chemical synthesis.
Traditional methods of enzyme engineering, such as directed evolution, involve randomly mutating proteins and evaluating the outcomes in search of desired traits. This iterative approach can be slow and resource-intensive, as scientists often have to perform numerous rounds of mutations to achieve an enzyme with the desired function. Dong envisions a shift toward “inverse design” aided by machine learning. Rather than beginning with random mutations, her approach starts with a target functionality. By combining AI with physics-based simulations, her team can predict which protein sequences will produce the desired reaction characteristics, effectively guiding protein engineering in a more focused and cost-effective manner.
Artificial intelligence, paired with advanced simulations, also allows for the prediction of enzyme properties following specific mutations, which reduces the need for costly experimental rounds in the directed evolution process. This framework has the potential to predict functional outcomes without the extensive trial-and-error traditionally required, offering a faster, more affordable path to designing optimized enzymes. “With this physics-based approach, you’re more likely to find the protein variant you need in a fraction of the time and cost,” says Dong.
In 2024, the Nobel Prize in Chemistry was partly awarded to a pioneer in computational protein design, underscoring the relevance and promise of Dong’s work. By pushing the boundaries of protein engineering into the realm of computational design, her team is at the forefront of a transformative approach that could redefine how we discover and synthesize new chemicals. Dong sees the potential for her computational framework to be a game-changer, one that might ultimately revolutionize not only chemical discovery but also the entire field of catalysis, bringing complex new functionalities to enzymes and opening up unprecedented possibilities for sustainable, efficient chemical processes.
Source: Northeastern University