Increasing batch-to-batch reproducibility of CHO-cell cultures using a model predictive control approach.

Area of Science:

• Biotechnology
• Bioprocess Engineering
• Cell Culture Technology

Background:

• Recombinant therapeutic protein production relies on consistent animal cell culture performance.
• Chinese hamster ovary (CHO) cells are widely used for producing proteins like erythropoietin (EPO).
• Achieving high batch-to-batch reproducibility in cell culture is critical for reliable therapeutic protein manufacturing.

Purpose of the Study:

• To implement a model predictive control (MPC) strategy for enhancing batch-to-batch reproducibility in CHO cell cultures.
• To optimize specific growth rate considering productivity, protein quality, and process controllability.
• To indirectly control cell growth and concentration by manipulating oxygen consumption via glutamine feed.

Main Methods:

• Utilized a model predictive control (MPC) strategy incorporating a classical process model.
• Controlled oxygen mass consumption as a proxy for specific biomass growth rate and cell concentration.
• Manipulated glutamine feed rate to manage oxygen consumption.
• Adapted model parameters during cultivation to account for dynamic process changes.

Main Results:

• Achieved high batch-to-batch reproducibility in CHO cell cultures for EPO production.
• Demonstrated effective control of oxygen consumption profiles, leading to consistent cell and protein titers.
• Observed minimal deviations in optimal operational trajectories due to predictive capabilities.
• Maintained consistent sialylation patterns across all cultivation runs.

Conclusions:

• Model predictive control is a viable strategy for ensuring reproducibility in biopharmaceutical manufacturing.
• Controlling oxygen consumption via feed manipulation provides a robust method for managing cell growth and productivity.
• Adaptive model parameterization enhances the robustness of MPC in dynamic bioprocesses.