Continuous Enzyme Evolution – solving bottlenecks in enzyme engineering to design next-generation biocatalysts

Introduction

Directed evolution has revolutionized the application of enzymes in industrial settings by allowing users to tailor the properties and activities of biocatalysts to their needs. But classic directed evolution is notoriously labor- and time-intensive, as it manually stages mutation, selection, and amplification cycles.

Continuous Evolution

In contrast, continuous evolution (CE) approaches aim to achieve these steps within a replicating organism, making it possible to engineer efficient enzymes in a matter of days rather than months or years. Unfortunately, current CE approaches are typically applicable only to model enzymes with little industrial value.

Project Goals

To unleash the full potential of CE, we will develop a scalable, low-tech CE platform, which is readily applicable to biocatalysts that provide value-added products. Toward this end, we will:

1. Merge a versatile selection system we recently developed with strategies to diversify the genes of targeted enzymes in vivo.
2. Implement an autonomous setup to grow bacterial populations continuously.

Combined, the resulting CE platform will enable us to engineer biocatalysts along many and long evolutionary trajectories.

Analysis and Machine Learning

Moreover, analyzing the fate of these populations by sequencing will allow us to map the sequence-structure-function relationships of these biocatalysts. Based on the systematic datasets generated in these efforts, we will train machine-learning (ML) models to predict functional sequences.

Design-Build-Test-Learn Cycle

Lastly, in a ML-directed CE approach, we will establish a design-build-test-learn cycle to improve models and guide CEs toward promising, but otherwise inaccessible sequence spaces.

Conclusion

Overall, ContiZymes will overcome unaddressed challenges associated with the application of biocatalysts that promote sought-after C-C, C-halogen, and C-N-bond forming reactions. We will not only engineer these enzymes at an unprecedented rate and scale, but also map their fitness landscapes and take a critical step toward the on-demand design of next-generation biocatalysts.