Real‐time amino acid and glucose monitoring system for the automatic control of nutrient feeding in CHO cell culture using Raman spectroscopy

Abstract

An innovative, Raman spectroscopy‐based monitoring and control system is introduced in this paper for designing dynamic feeding strategies that allow the maintenance of key cellular nutrients at an ideal level in Chinese hamster ovary cell culture. The Partial Least Squares calibration models built for glucose, lactate and 16 (out of 20) individual amino acids had very good predictive power with low root mean square errors values and high square correlation coefficients. The developed models used for real‐time measurement of nutrient and by‐product concentrations allowed us to gain better insight into the metabolic behavior and nutritional consumption of cells. To establish a more beneficial nutritional environment for the cells, two types of dynamic feeding strategies were used to control the delivery of two‐part multi‐component feed media according to the prediction of Raman models (glucose or arginine). As a result, instead of high fluctuations, the nutrients (glucose together with amino acids) were maintained at the desired level providing a more balanced environment for the cells. Moreover, the use of amino acid‐based feeding control enabled to prevent the excessive nutrient replenishment and was economically beneficial by significantly reducing the amount of supplied feed medium compared to the glucose‐based dynamic fed culture.

Progress in fed-batch culture for recombinant protein production in CHO cells

“…Raman spectroscopy is used for monitoring nutrient and byproducts (glucose, lactate, and amino acid) (Domján et al 2022 ). Dynamic feeding strategies are used to control the delivery of feed medium via Raman spectroscopy.…”

“…In an early phase of the development, Domján et al established a partial least squares (PLS) calibration model that took the cell metabolic behavior and nutritional consumption into account. Subsequently, the nutrients are maintained at the desired level (Domján et al 2022 ). For mAb production in CHO cells, Eyster et al developed Raman spectroscopy to control lactate and glucose in fed-batch culture, and established PLS model for predicting glucose and lactate concentrations.…”

Advances in bioreactor control for production of biotherapeutic products

“…Deployment of MVDA and machine learning (ML) algorithms has not been approved in entirety by the regulators due to limited experience that the industry has with respect to their commercial deployment (Shah et al, 2019). In recent years, regulatory authorities ILLUSTRATION 2 Case Study-Raman spectroscopy-based monitoring and control system (Domján et al, 2022) Raman spectroscopy-based monitoring and control system was developed for deciding the feeding strategy for a Chinese Hamster Ovary (CHO) cell culture. Partial least square model was built for glucose, lactate, and amino acids concentration.…”

1H NMR‐based process understanding and biochemical marker identification methodology for monitoring CHO cell culture process during commercial‐scale manufacturing

“…Therefore, the usage of applicable tools to monitor critical biochemical indicators for cell culture has turned out to be a hot study subject among biopharmaceutical enterprises in the latest years. In previous studies, the monitoring and control of the cell culture processes mainly focused on physiological conditions and several significant biochemical indexes such as concentration of critical metabolites, and viable cell density (VCD), total cell density (TCD) [ 9–13 ] …”

“…In previous studies, the monitoring and control of the cell culture processes mainly focused on physiological conditions and several significant biochemical indexes such as concentration of critical metabolites, and viable cell density (VCD), total cell density (TCD). [9][10][11][12][13] Among all of the process analytical methods, proton nuclear magnetic resonance ( 1 H NMR) based metabolomics has been taken as a conventional fermentation process monitoring tool. 1 H NMR has been applied for predicting the bioprocess performance, and for early identification of potential biochemical markers linked to cell starvation, viability, and expansion, to favor the timely control of culture conditions, through global multi-component analysis.…”