Exploring the overproduction of amino acids using the bilevel optimization framework OptKnock

Area of Science:

• Metabolic Engineering
• Systems Biology
• Computational Biology

Background:

• The metabolic network of Escherichia coli is complex, requiring strategic modifications for targeted metabolite overproduction.
• Identifying optimal gene knockouts for enhancing amino acid production is crucial for biotechnology.

Purpose of the Study:

• To modify and extend the OptKnock bilevel optimization framework for improved gene knockout strategy identification.
• To identify strategies for overproducing amino acids and their precursors in Escherichia coli.

Main Methods:

• Extended the OptKnock framework to include gene deletions, transport rates (CO2, NH3, O2), and secretion pathways as optimization variables.
• Applied the modified framework to the Escherichia coli metabolic network to predict gene knockout strategies.
• Analyzed computational results to identify key pathways and non-intuitive gene targets for metabolite overproduction.

Main Results:

• Demonstrated the importance of manipulating energy pathways, controlling nitrogen/oxygen uptake, and blocking competing secretion pathways.
• Identified specific non-intuitive gene knockouts, such as pyruvate kinase, phosphotransacetylase, and ATPase, for glutamate overproduction.
• Found that phosphofructokinase and ATPase knockouts, alongside others, can enhance alanine yield.

Conclusions:

• The enhanced OptKnock framework provides valuable, often non-intuitive, gene knockout suggestions for amino acid overproduction.
• The study highlights the significance of targeting central metabolism, energy pathways, and transport/secretion mechanisms.
• The framework offers a foundation for incorporating additional modeling refinements, such as regulatory constraints, in future studies.