Development and Analysis of a Mathematical Model for Antibody-Producing GS-NS0 Cells Under Normal and Hyperosmotic Culture Conditions

Abstract

The GS-NS0 cell line is industrially important and is currently used for the large-scale production of several therapeutic monoclonal antibodies. A novel hybrid model, consisting of both unstructured and structured elements, has been developed to describe cell growth and death, metabolism, and antibody production in the GS-NS0 system under normal culture conditions. A comparison between the hybrid model and a large-scale single-cell model (SCM) describing detailed metabolic processes verified the predictive ability of the hybrid model (when compared with experimental data) and highlighted the practical difficulties involved in utilizing complex models. Global sensitivity analysis (GSA) on the hybrid model identified the specific transcription and translation rates of heavy and light immunoglobulin chains as parameters with the largest impact on the antibody production process. This information, together with the addition of a 24-h lag phase, resulted in the successful extension of the hybrid model to represent GS-NS0 system behavior under hyperosmotic culture conditions.

Cell line‐specific control of recombinant monoclonal antibody production by CHO cells

“…The mathematical model consists of a set of ordinary differential equations (Supplementary Material 2) and is based on a dynamic model of Mab synthesis within GS-NS0 cells published by Ho et al (2006). The model was built and solved using Sentero, an in-house graphical software tool for biochemical modeling (Wilkinson et al, 2008) that uses Matlab as the simulation engine.…”

“…In an earlier model of Mab synthesis within GS-NS0 cells developed by Ho et al (2006) the pathway proceeds via a half Mab intermediate as proposed by Percy et al (1975). To test this and ensure that we employed the correct IgG 4 assembly sequence for our mathematical model we investigated the sequence and kinetics of the Mab assembly pathway in vivo by means of a standard radiolabeling experiment, utilizing both cell lines 47 and 38, where intracellular Mab synthetic intermediates were purified from cell lysates using a Protein A affinity purification step.…”

“…Two more recent papers have developed more advanced mathematical tools to understand the control structure of the Mab biosynthetic pathway. Gonzalez et al (2001) developed a generically applicable metabolic control analysis model of Mab biosynthesis within eukaryotic cells, and Mab production within GS-NS0 cells has recently been modeled by Ho et al (2006). In this model HC and LC transcription and translation were identified as critical parameters controlling productivity throughout culture.…”

Microarray profiling of preselected CHO host cell subclones identifies gene expression patterns associated with in‐creased production capacity

“…Overall, the focus on fold changes distracts attention from changes in gene expression that are regulated at a more physiologic scale such as for instance transcription factors that have an effect on the cell upon slight changes in their expression [30]: if within a pathway each individual gene is upregulated 1.4-fold, none of these will show up in the list of differentially expressed genes when using a threshold of 1.5, however, the overall flux through the pathway will increase significantly. [7,66,69,70] Down regulation of growth [7,32,66,68] In case of antibodies: enhanced [4,69,71,[76][77][78]] assembly and secretion due to a surplus of light chain over heavy chain These challenges in identifying relevant genes and understanding their biological roles are further compounded by the fact that many genes have functional assignments in the GO or KEGG annotation databases that reflect the context in which they were first discovered, which may be completely irrelevant to cells grown in suspension in a protein free medium (such as organism development, tissue specific functions and similar). Moreover, pleiotropy, the phenomenon where a single gene may have multiple functions, which may not all be known [31], can make the interpretation of results difficult.…”

Dynamic Optimization of Bioprocesses

“…The model presented herein can be extended to describe other important process variables that can act as controls. Examples include cell culture temperature, which can enhance the protein productivity [2,6], culture pH, which can affect cell growth and viability [7], and medium osmolarity, which can influence specific antibody production rates [8]. …”