Generation and Use of a Genome-Scale Model to Describe Metabolic Shift in CHO Cells

Mammalian cell-based processes have become one of the main approaches for production of biopharmaceuticals. The increasing demand for these products has been a driving force for the development of new strategies to improve cellular productivity in culture. Most of these approaches are focused on central carbon metabolism, particularly on the reduction of flux through glycolysis, or reversion of lactate dehydrogenase activity towards reduction of lactate production (metabolic shift) since high concentrations of this metabolite are known to be detrimental for cellular growth and product synthesis.

Understanding the mechanisms that give rise to these cellular processes are relevant for the development of new cell lines and cell culture processes. In order to achieve this goal genome-scale models have been used to represent cellular metabolism, including genomic information, chemical reactions and the relationship between them. However, the use of genome-based models for mammalian cells has been sparse due to the inherent complexity of these systems and the lack of specific genomic data for these cell lines.

In this work we developed a genome-scale model based on specific genomic information for CHO cells provided by the CHO genome initiative. To achieve this goal we used Pantograph, a tool that automatically generates genome-scale models for eukaryotic organisms. This tool requires a template model and annotated genome for both template and the objective organism. The relationship between model and objective organisms is obtained by ortholog search, which allows identifying conserved biochemical functions from the template organism. 

Two approaches were used for defining a template model: the first one considered a Mus musculus metabolic reconstruction based on the human genome-scale model Recon 1; the second one uses an improved version of a published Hybridoma model based on Mus musculus information. These reconstructions were complemented with ortholog information for Mus musculus, Homo sapiens and CHO to generate a genome-scale model for this cell line.

The obtained model has 1,041 genes, 3,459 reactions distributed among cytosol, mitochondria, golgi, lysosome, ribosome, peroxysome and nucleus. Validation was performed using specific experimental information forgrowth on alternative carbon sources and gene knockouts on CHO cells.

This is the first genome- scale model developed for CHO cells that has been used to study the metabolic shift phenomenon. The model exhibits the potential to represent the altered states of cellular metabolism observed experimentally by integrating reported changes on gene expression for this state, where a shift from lactate production to consumption can be associated to positive effects in the culture in terms of cell growth and productivity.