Pathway Evolution Through a Bottlenecking-Debottlenecking Strategy and Machine Learning-Aided Flux Balancing.

Area of Science:

• Metabolic Engineering
• Synthetic Biology
• Biocatalysis

Background:

• Enzyme evolution is key for producing valuable chemicals but is often hindered by complex genetic interactions (epistasis).
• The naringenin biosynthetic pathway exhibits complex epistasis, complicating traditional directed evolution methods.

Purpose of the Study:

• To develop a robust biofoundry-assisted strategy for predictable, parallel evolution of pathway enzymes.
• To optimize metabolic pathways using machine learning for enhanced chemical production.
• To demonstrate the broad applicability of the strategy for automated chassis construction.

Main Methods:

• A biofoundry approach was used for parallel enzyme evolution and pathway balancing.
• Machine learning model (ProEnsemble) optimized gene transcription for pathway balance.
• Engineered Escherichia coli chassis constructed with evolved and balanced pathway genes.

Main Results:

• Achieved a predictable evolutionary trajectory for pathway enzymes within six weeks.
• Generated 3.65 g/L of naringenin, a high-value chemical, in the engineered E. coli.
• Demonstrated enhanced production of other flavonoids using the optimized naringenin chassis.

Conclusions:

• The developed biofoundry strategy enables efficient and predictable evolution of metabolic pathways.
• This approach facilitates automated chassis construction for diverse chemical biosynthesis applications.
• The strategy is adaptable for various enzymes and metabolic pathways, advancing synthetic biology.