UC San Diego

Chinese hamster ovary (CHO) cells are the dominant host for producing biotherapeutic proteins. Despite a long history of use, the techniques used to generate high quantities of recombinant drugs from this cell line remain largely unchanged. This is—at least partly—due to the complexity of mammalian cell lines. Historically—in contrast to microbial systems—CHO cells have been viewed as a bioprocessing black box, a system to optimize around rather than engineer. In this dissertation, advancements in gene editing tools and the availability of a genomic sequence are used to establish a framework to move away from that perspective and enable rational engineering of CHO cells for desirable phenotypes. Specifically, a genome-scale model of CHO cell metabolism, iCHO1766, was reconstructed from the genome sequence. iCHO1766 contains the biochemical basis for growth and protein production in CHO cells and predicts growth rates as well as CHO-specific amino acid auxotrophies. This model was also deployed to identify strategies that efficiently redirect resources from growth to protein production. Using the recently characterized CRISPR-Cas9 system, we were able to eliminate the Warburg effect in CHO cells by simultaneously knocking out lactate dehydrogenase and regulators involved in a negative feedback loop that typically inhibits pyruvate conversion to acetyl-CoA. In contrast to long-standing assumptions about the role of aerobic glycolysis, Warburg-null cells maintain the same growth rate as wildtype cells while consuming less glucose without increasing oxygen uptake to compensate for lost glycolytic ATP. The cells produce negligible lactate—allowing prolonged growth to higher cell densities—and remain viable for generating recombinant protein producing cell lines. Introducing this phenotype into a CHO cell line already producing a biotherapeutic antibody maintained protein production and improved glycan galactosylation. Thus, by leveraging accessible gene editing techniques, an improved CHO cell line has been generated. With newly available systems biology models, the tools for further rational engineering toward better and more affordable biotherapeutics are now available.