Using Cutting-Edge Tools to Optimize Cell-Culture Media Development

Yaron Silberberg

September 28, 2022

3 Min Read

Although culture-media optimization accounts for a relatively small part of process development, selections made at that stage strongly influence overall bioprocess productivity. Yaron Silberberg (chief scientist at the Ajinomoto Genexine CELLiST Solution Center) joined BPI in July 2022 to present strategies for enhancing media performance. He focused on how emerging tools for statistical analysis and process control facilitate media development for cultures based on Chinese hamster ovary (CHO) cell lines.

Silberberg’s Presentation
Because materials applied during cell-line development are not optimized for large-scale protein production, host cells usually undergo sequential adaptation to a medium suitable for industrial applications. Batch and/or fed-batch cultures in shake flasks or microbioreactors are used to screen potential feed regimens. Then, optimization activities are conducted to assess the effects of process parameters and supplements on expression titers and batch productivity. Advanced optimization steps study influences of components on protein quality.

Emerging solutions for statistical analysis can facilitate optimization, Silberberg said. A culture medium can contain >50 components, making it impractical to screen synergies across all components in search of an ideal formulation. Ajinomoto Genexine leverages design of experiments (DoE) and computational modeling to identify media components that limit or enhance culture performance.

Silberberg’s unit used such a strategy to optimize the amino-acid (AA) composition of a customer’s medium. The team used a Plackett–Burman DoE approach to study how increases and decreases in 22 AAs influenced culture productivity. Results identified four critical components, which became the parameters for a full-factorial DoE study. Silberberg’s team created 16 AA formulations to supplement fed-batch cultures grown in microbioreactors. Spent-media samples were analyzed. Finally, the team used JMP software to identify synergistic effects between AAs, enabling development of an optimized supplementation strategy.

Multiomic tools can enhance media screening further. Silberberg described his company’s partnership with Sungkyunkwan University (SKKU) in South Korea to establish a framework for model-guided media development. The workflow presented began with data collection from cultures of two CHO-K1 cell lines grown in two media. Those findings underwent multivariate data analysis to identify correlations between media components and culture productivity. Components were assigned variable importance of projection (VIP) scores based on their relevance to critical metabolic pathways, and materials with high VIP scores were studied further using genome-level mechanistic models.

Models showed that one of the two media had limited access to flavin adenine dinucleotide (FAD) in the electron transport chain. Subsequent analysis identified coenzyme Q10 (CoQ10) as a critical component in that pathway. Silberberg’s team adjusted the formulation to increase the CoQ10 dose. In subsequent wet experiments, the optimized medium showed significantly higher productivity.

Ajinomoto Genexine has partnered with Insilico Biotechnology to help customers enhance feed protocols using digital twins. Customers provide Ajinomoto Genexine with a cell line for fed-batch studies. Resulting data are entered into an Insilico Biotechnology model to screen dynamic feeding strategies. Alternatively, customers can provide spent-media data for modeling. Feed strategies often can be developed in one to two months, including time for wet and in silico experiments. Silberberg highlighted that a recent customer increased the productivity of a previously optimized culture process by >25% through implementation of a model-identified dynamic feed strategy.

Ajinomoto Genexine continues to develop its CELLiST product lineup and customer services to support media optimization. New offerings include novel cysteine (Cys) supplements that are designed to inhibit oxidation and increase Cys supply to cultured cells.

Questions and Answers
When can a different brand of culture media be introduced? A medium can be changed any time during process development, even during clinical studies. Depending on the stage, users might need to submit change notifications to regulatory agencies.

How important are supplements such as glutamine to culture performance? Although results differ across hosts and products, glutamine supplementation often increases productivity, especially while cells are adapting to new media. Ajinomoto offers dipeptide formulations that are designed for enhanced stability.

What titers have Ajinomoto Genexine customers been able to achieve after optimization? Results are product specific, but processes often yield >8 g/L, and many low-producing cell lines have increased in productivity two- or threefold.

Find the full webinar online at

You May Also Like