Interplay of stability and dynamics in the optimization of a highly proficient de novo enzyme

For decades, scientists have dreamed of building life's machinery from scratch. Nature spends billions of years evolving proteins, the molecular tools that drive every chemical reaction inside a living cell, to perform specific tasks with incredible speed and precision. Researchers now ask if they can skip the evolutionary wait and design these tools directly on a computer. This field, known as de novo enzyme design, aims to create proteins that can perform chemical reactions that do not exist in nature. The goal is not just to make a shape that fits a molecule, but to build a machine that can grab a raw material, hold it in the perfect position, and force it to change into something new. While creating a protein that works at all is a major achievement, making one that works as well as the enzymes found in nature remains a profound challenge.

A team of researchers recently took a deep dive into this challenge by studying a specific computer-designed enzyme that they improved over time. They started with a version created entirely by algorithms, which was then refined through two rounds of directed evolution, a process where scientists mimic natural selection in the lab to find better-performing versions. By combining computer modeling with physical experiments, they traced the journey from a rough digital sketch to a highly efficient catalyst. Their work reveals that the secret to a powerful designed enzyme lies in a delicate balance between two opposing forces: the need for the protein to stay rigid and stable, and the need for certain parts to remain flexible enough to move.

The original computer design focused almost exclusively on how the enzyme would hold the transition state, which is the fleeting, high-energy moment when a chemical reaction is actually happening. The scientists built the protein to lock this moment in place, but they did not design it to hold the starting material, known as the substrate, effectively. As a result, the initial enzyme struggled to grab its target. Through directed evolution, the researchers introduced mutations that improved how the enzyme held the substrate, creating a more stable starting point for the reaction. However, the most surprising discovery came when they looked at the final, most efficient version of the enzyme.

When the researchers examined the structure of this optimized enzyme, they found that binding a molecule that mimics the transition state caused the entire protein backbone to become more stable and rigid. This widespread stiffening suggests that the whole protein acts as a supportive scaffold, holding the active site in the precise alignment needed for the reaction to occur. Yet, there was a crucial exception. Near the entrance to the active site, a specific section of the protein containing a helix did not become rigid. Instead, the introduction of specific amino acids, glycine and proline, made this area more dynamic and flexible. This flexibility turned out to be essential for the enzyme's high activity, allowing it to function efficiently while the rest of the structure remained locked in place.

The study demonstrates that creating a highly proficient enzyme requires optimizing multiple features simultaneously. It is not enough to simply design a pocket that fits the transition state; the entire protein must contribute to the process. The initial design failed because it ignored the need for a stable starting complex, while the final success came from a combination of global stability and local flexibility. The researchers found that the entire protein contributes to catalysis in the most optimized enzyme, showing that the path to high activity involves tuning the whole system, not just the active center. This work highlights the complex interplay between stability and movement that nature has perfected over eons and that scientists are only beginning to replicate in the lab.