Unlocking the Potential of CHO Cells: Advancing Bioproduction with Next-Generation Vector Design

Traditional cell line development to support bioprocessing often has relied on a “one-size-fits-all” vector, which can lead to suboptimal expression levels in Chinese hamster ovary (CHO) cell cultures. Scott Estes (head of cell line development at Asimov) and his team have brought a rational approach to how cells are engineered. The CHO Edge system integrates knowledge from experts in cell line engineering, synthetic biology, software design, machine learning, and omics-driven modeling of metabolism and gene expression.

Estes’s Presentation
Estes explained that the CHO Edge platform engine is based upon Asimov’s robust CHO-K1 glutamine synthetase (GS) knock-out line. Asimov adapted CHO-K1 host cells from the European Collection of Authenticated Cell Cultures (ECACC) and used them to create glutamine auxotrophs. After genotypic and the phenotypic levels were characterized, a cell line was sent to Charles River Laboratories for creation of a master cell bank and safety, sterility, and identity testing. Asimov now licenses CHO Edge technology to customers and also offers it as a service.

Estes explained that the CHO Edge platform has an extensive library of genetic parts that include both natural sequences and thousands of synthetic elements that are characterized for performance within Asimov’s CHO-K1 host. Thus, Estes said, the platform broadens the industry’s vector toolbox and enables scientists to explore different expression strategies, which can greatly enhance cell-line productivity. Asimov integrates a genetic payload into its host cells by using a proprietary hyperactive transposase. Expression vectors (mRNA encoding the transposase and gene of interest) are delivered by electroporation. The process can integrate multiple copies in stable transcription, enabling users to achieve productive and robust cell lines.

Kernel computer-aided design software serves as a repository for Asimov’s genetic-parts library. Clients can use the platform to interface with the library and build permutations of expression vectors with different types of elements. The software contains proprietary algorithms for codon optimization and signal-peptide prediction. The algorithms can explore range-of-chain ratios and degrees of attenuation for selectable markers, ensuring maximum productivity for recombinant proteins.

Scientists can use characterized elements to identify vector permutations. Estes described a case study in which his team produced four different antibodies in six expression vectors. For the first antibody, productivity was maximized with a balanced heavy-to-light chain ratio. The second and third antibodies benefited from an overexpression of the light chain relative to the heavy chain. For the fourth antibody the team tested different vector strategies and varied the design to achieve a pool titer >6 g/L with no cloning out. Estes reiterated that by expanding the design space, his team could improve vector productivity.

In another set of experiments, Asimov explored a broad vector design space with 14 unique constructs. Of the two cell populations that emerged, one produced high titers but with a low percentage of desired heterodimer. The second showed slightly lower titers but a much higher percentage of desired product. Estes said that when such pools are cloned out, users can obtain a favorable titer of 8–11 g/L across the top 12 clones with 80–90% of the final titer being composed of the desired product.

Omics-driven systems further enhance process modeling. Asimov has generated an immense amount of data to feed its expression and cell-metabolism models. Kernel software uses such data to further train models and make them predictive. Asimov’s CHO-K1–based system consistently delivers 6–10 g/L and remains stable over many generations, removing that parameter from the critical path in cell line development. To learn more, visit www.asimov.com/CHO.

Questions and Answers
How many different vector designs do you include in a cell line development project? That depends on the complexity of the molecule being expressed. For a straightforward immunoglobin G (IgG), we use about six vector permutations to explore the chain ratios and GS attenuation. However, when we start moving into constructs of greater complexity and more chains, we broaden that vector design space with 12 to 16 different vectors to maximize our probability of success.

How do you assess monoclonality of cell lines? We use redundancy in assuring monoclonality. First, we use an F.SIGHT cell printer, which is an instrument that dispenses single cells into wells for imaging. We follow up that assessment with a Solentim Cell Metric plate scanner (Advanced Instruments), which images single cells as they are dispensed and when they are in their respective wells.

Find the full webinar online at www.bioprocessintl.com/category/webinars.