Using highly efficient nonlinear experimental design methods for optimization of Lactococcus lactis fermentation in...

Area of Science:

• Biotechnology
• Microbial Fermentation
• Process Optimization

Background:

• Fermentation media and process optimization are challenging due to high dimensionality and nonlinearity.
• Developing efficient optimization strategies is crucial for microbial growth and production.

Purpose of the Study:

• To develop and evaluate two novel, efficient methods for experimental fermentation optimization.
• To optimize conditions for maximum cell growth of Lactococcus lactis IL1403 using a chemically defined medium.

Main Methods:

• A truncated genetic algorithm combined with a neural network model for experiment selection.
• Information theory coupled with Bayesian regularized neural networks for experiment selection.
• Experimental validation using Lactococcus lactis IL1403 with 19 defined media components, temperature, and initial pH.

Main Results:

• Novel methods achieved comparable or higher maximum cell growth than traditional statistical methods.
• Optimization was achieved in approximately half the number of experiments (73-94 vs. 161).
• Developed chemically defined media supported 3.5-4 times higher cell density than synthetic media and 72% higher than complex M17 medium.

Conclusions:

• The novel optimization methods are highly efficient for developing fermentation processes.
• The developed chemically defined media support significantly enhanced cell growth for Lactococcus lactis IL1403.
• These findings offer improved strategies for microbial fermentation optimization and media development.