Strangers in a foreign land: 'Yeastizing' plant enzymes

Area of Science:

• Synthetic biology
• Biotechnology
• Metabolic engineering

Background:

• Microbial platforms, particularly yeast, are cost-effective for producing plant compounds.
• Plant enzymes often fail in yeast due to poor cellular adaptation.
• Current enzyme optimization methods are labor-intensive and enzyme-specific.

Purpose of the Study:

• To propose a computational pipeline for redesigning plant enzymes for yeast environments.
• To overcome limitations of current labor-intensive enzyme engineering approaches.
• To enable efficient establishment of plant metabolic pathways in yeast.

Main Methods:

• Summarizing current engineering approaches for plant enzyme optimization in yeast.
• Analyzing sequence features distinguishing plant and yeast enzymes.
• Developing a data-driven machine learning framework to extract 'yeastizing' rules.
• Proposing integration into a design-build-test cycle.

Main Results:

• Plant and yeast enzymes exhibit generalizable, distinct sequence features.
• A machine learning framework can extract rules for adapting plant enzymes to yeast.
• The proposed computational pipeline offers a broadly applicable solution.

Conclusions:

• Computational redesign of plant enzymes can improve yeast compatibility.
• Machine learning provides a scalable method for enzyme adaptation.
• This approach facilitates the development of yeast-based cellular factories for plant compound production.