Enzyme engineering plays a central role in developing efficient biocatalysts for biotechnology, biomedicine, and life sciences. Apart from classical rational design and directed evolution approaches, machine learning methods have been increasingly applied to find patterns in data that help predict protein structures, improve enzyme stability, solubility, and function, predict substrate specificity, and guide rational protein design. In this Perspective, we analyze the state of the art in databases and methods used for training and validating predictors in enzyme engineering. We discuss current limitations and challenges which the community is facing and recent advancements in experimental and theoretical methods that have the potential to address those challenges.
Whether designed computationally or uncovered in activity screening, enzymes repurposed for biocatalysis rarely start at the peak of proficiency. However, directed evolution can in some cases increase catalytic efficiency of a poor enzyme by many orders of magnitude. Otten et al. used a suite of biochemical techniques to investigate the origins of rate enhancement in a previously evolved model enzyme. Two conformational states are present in the initial, computationally designed enzyme, but only one is active. Shifting the population toward the active state is one factor in increasing catalytic efficiency during evolution. Single mutations do not greatly increase activity, but the synergistic combination of just two out of 17 substitutions can provide most of the rate enhancement seen in the final, evolved enzyme. Science , this issue p.  The advent of biocatalysts designed computationally and optimized by laboratory evolution provides an opportunity to explore molecular strategies for augmenting catalytic function. Applying a suite of nuclear magnetic resonance, crystallography, and stopped-flow techniques to an enzyme designed for an elementary proton transfer reaction, we show how directed evolution gradually altered the conformational ensemble of the protein scaffold to populate a narrow, highly active conformational ensemble and accelerate this transformation by nearly nine orders of magnitude. Mutations acquired during optimization enabled global conformational changes, including high-energy backbone rearrangements, that cooperatively organized the catalytic base and oxyanion stabilizer, thus perfecting transition-state stabilization. The development of protein catalysts for many chemical transformations could be facilitated by explicitly sampling conformational substates during design and specifically stabilizing productive substates over all unproductive conformations. : /lookup/doi/10.1126/science.abd3623
In a major breakthrough toward medical'magic mushrooms,' scientists have unraveled the enzymes behind the ingredient responsible for their psychedelic effect. Research over the last few decades has suggested that the compound psilocybin may have a number of therapeutic benefits, with potential to help treat anxiety, depression, and even addiction. In a new study, scientists have characterized the four enzymes mushrooms use to make this compound for the first time, setting the stage for pharmaceutical production of the'powerful psychedelic fungal drug.' Research over the last few decades has suggested that the compound psilocybin may have a number of therapeutic benefits, with potential to help treat anxiety, depression, and even addiction. But until now, the'recipe' for psilocybin has remained a mystery Magic mushrooms are the safest recreational drug to take, researchers said on the back of a new international survey in May. Just one in 500 people will be taken to hospital on the back of dodgy side-effects of the hallucinogenic, the study suggested.
Novel proteins, created from scratch with no particular design in mind, can sometimes do the work of a natural protein. The discovery may widen the toolkit of synthetic biologists trying to build bespoke organisms. There are more proteins possible than there are atoms in the universe, and yet evolution has tested only a minuscule fraction of them. No one knows whether the vast, untried space of proteins includes some that could have biological uses. Until now, most researchers assembling novel proteins have meticulously selected each amino acid building block so that the resulting protein folds precisely into a pre-planned shape that closely fits the molecules it is intended to interact with.
A team lead by researchers in the Pritzker School of Molecular Engineering (PME) at the University of Chicago reports that it has developed an artificial intelligence-led process that uses big data to design new proteins that could have implications across the healthcare, agriculture, and energy sectors. By developing machine-learning models that can review protein information culled from genome databases, the scientists say they found relatively simple design rules for building artificial proteins. When the team constructed these artificial proteins in the lab, they discovered that they performed chemistries so well that they rivaled those found in nature. "We have all wondered how a simple process like evolution can lead to such a high-performance material as a protein," said Rama Ranganathan, PhD, Joseph Regenstein Professor in the Department of Biochemistry and Molecular Biology, Pritzker Molecular Engineering, and the College. "We found that genome data contains enormous amounts of information about the basic rules of protein structure and function, and now we've been able to bottle nature's rules to create proteins ourselves."