Improving CHO cell line performance through cell and bioprocess engineering
A continuing goal in mammalian biomanufacturing processes is to evaluate, modify, and ultimately enhance the performance of CHO cells in culture. Indeed, one of the major challenges in CHO cell culture is long term cell line stability. Torres et al . took a deeper look into this issue, examining changes in cell culture performance, gene expression, and metabolism over time for two cell lines. Especially interesting was the upregulation of genes involved in cell proliferation and survival concomitantly with changes in metabolites’ uptake and production rates. Acknowledging the impact that culture time has on the cell, researchers are now exploring ways to improve cell line stability, aiming to design more predictable, consistent, and productive expression system. One of such approaches is targeted integration; however, identifying a CHO cell genomic loci capable of supporting high-level protein expression is still a bottleneck. To address this need, Leeet al ., implemented the “Thousands of Reporters Integrated in Parallel” (TRIP) high-throughput screening method, identifying several hotspot candidates in the CHO genome exhibiting high transgene mRNA expression. In another study, Marx et al. presented a fast and robust method -the nanopore Cas9-targeted sequencing (nCats) pipeline - to characterize cell clones and isolate the most promising ones. This method was able to identify integration sites, the composition of the integrated sequence, and the DNA methylation status in CHO cells in a single sequencing run. Building up on CRISPR/Cas9 technology for mammalian cells, Lee et al . developed an all-in one reporter system to quantify gene disruption and site-specific integration (SSI) in CHO cells. Using this system, it was possible to identify specific molecules (inhibitors of DNA repair pathways) that enhance SSI efficiency and thus accelerate cell engineering. Another approach to further improve mammalian cell factories is to ameliorate cell’s capacity to handle proteotoxic stress, which can result in cellular apoptosis. In Segatori et al. , the challenges and opportunities in synthetic biology for improving these programmable cell factories is detailed. An important contribution to the field of cell line and metabolic engineering is provided by Kontoravdi et al. with a review on the current state of CHO genome-scale metabolic models (GEM), their inability to model intracellular metabolism and capture extracellular phenotypes. In addition, Kontoravdi et al . presented an improved GEM, iCHO2441, as well as two cell line specific GEMs for CHO-S and CHO-K1 that may serve as foundations for better design and assess next-generation flux analysis techniques. Jimenez del Val et al . propose a more compact network model, CHOmpact, that can provide improved interpretations from simulations, including identification of shifts in key metabolic behaviours. This model could also serve as a platform for dynamic models used for process control and optimization. The advancement in process analytical techniques (PAT) and artificial intelligence (AI) has enabled the generation of enormous culture datasets from biomanufacturing processes. AI-based data-driven models permit the correlation of biological and process conditions and cell culture states. This approach was exploited by Lee et al.that describe data-driven prediction models for forecasting multi-step ahead profiles of mAbs produced in CHO towards bioprocess digital twins.
A complementary approach to cell line engineering is the development of improved biomanufacturing processes. To address that need, Ben Yahiaet al. worked on intensified processes by optimizing the feeding strategy and specific power input (P/V) in a high-cell-density (HCD) seed bioreactor operated in fed-batch mode towards improved monoclonal antibodies (mAb) expression in the production bioreactor. Interestingly they report a positive impact of cellular “organized stress” in the seed bioreactor on the production performance. Chotteau et al . implemented a Design of Experiment (DoE) approach to design an optimal CHO culture medium capable of supporting the operation of microbioreactors in perfusion at HCD and low specific perfusion rates while maintaining constant specific product quality attributes such as N-glycosylation profile of the produced antibody. Using a similar statistical design methodology, Ladiwala et al . fine-tuned specific amino acids levels in reference basal and feed media in order to limit production of inhibitory metabolites, ultimately enhancing peak viable cell densities and product titers. Alternatively, Naik et al. looked at adding glycolysis inhibitors to limit the production of lactate as metabolic by-product in CHO cell cultures. They found that specific glucose analogs could lower peak lactate concentrations while also providing an increase in the final titers, although there was some changes in glycosylation patterns indicating an effect on product quality.