Dynamic Model Selection and Optimal Batch Design for Polyhydroxyalkanoate (PHA) Production by Cupriavidus necator.

Area of Science:

• Biotechnology and bioprocess engineering.
• Microbial metabolism and synthetic biology.

Background:

• Mathematical modeling is crucial for optimizing microbial polyhydroxyalkanoates (PHA) production, but model selection remains heuristic.
• Developing robust models requires accurate kinetic parameter identification and growth model selection strategies.

Purpose of the Study:

• To apply a parametric discretization approach for modeling PHA production in Cupriavidus necator.
• To establish reliable PHA production kinetics and select appropriate growth models from experimental data.
• To design an optimal batch startup policy for enhanced PHA accumulation.

Main Methods:

• Utilized a parametric discretization approach to model PHA production using sucrose and urea in Cupriavidus necator.
• Determined product formation kinetics from urea-free experiments and selected growth models from batch cultures with sucrose and urea.
• Employed logistic growth and Luedeking-Piret models, validated using R², adjusted R², AICc, cross-validation, confidence intervals, and sensitivity analysis.

Main Results:

• Selected logistic growth and Luedeking-Piret models with high goodness-of-fit (R²: 0.941, adjusted R²: 0.930, AICc: -42.764).
• Validated model fitness through cross-validation, confidence interval, and sensitivity analyses.
• Predicted an optimal batch startup policy yielding 2.030 g L⁻¹ PHA accumulation at 120 h.

Conclusions:

• The developed modeling framework effectively addresses over-parameterization and identifiability issues in PHA production models.
• This approach facilitates the design of optimal batch startup policies for microbial PHA bioproduction.
• The study provides a robust methodology for enhancing bioprocess design and PHA yield.