Understanding the Effects of Transgene Integration in CHO Cells using Targeted Locus Amplification (TLA) for Genomic Fingerprinting
April 24, 2022 Marie-Ange Kouassi
CHO cells are commonly used to produce biotherapeutics mostly by random integration of the desired expression cassette. This genetic modification of an organism inherent to genetic instability can lead to production as well as product instabilities and requires extensive clone screening and characterization campaigns. It is therefore important to identify the consequences of manipulation and integration events as early as possible during the cell line development process to prevent costly consequences.
To this day, there remains a limited understanding of random transgene insertion and its consequences in CHO (Chinese Hamster Ovary) cells. Although these cells are widely used in the production of biopharmaceuticals due to their robustness, high cell-specific productivity, and ability to express proteins with human-like structure, caution is required when handling them as they can be genetically unstable. This instability, as characterized for example by chromosomal instability, haploidy/polyploidy, or the loss of genes, can negatively impact productivity and stability causing lost resources and delayed timelines. Using Targeted Locus Amplification (TLA), a Next-Generation Sequencing (NGS) based technology, transgene insertion sites can be precisely determined following the random integration process. This can help to characterize the changes that occur to the surrounding genome and better predict its impact on clone performance. Stadermann et al. leveraged TLA to detect the genomic positions of transgene integration by generating genomic fingerprints of several CHO production cell lines.
After performing standard cell line development campaigns, specific clones were selected based on their optimal growth characteristics and high performance (titers of ≥1 and ≥3 g/L and specific productivities of ≥10 and ≥20 pg/cell for complex (mAb3) and IgG (mAb1, mAb2) modalities, respectively). Furthermore, the team investigated the impact of supertransfecting high-producing clones on integration integrity and clone performance.
The team then undertook TLA and processed the data in the Genedata Selector® software using the automated TLA workflow and reporting functionality. During this study, data analysis was also performed in the software to detect the specific integration sites and generate circular plots.
From the results, all high producers exhibited a single integration site, but with varying copies of the expression constructs. This suggested a concatenation of the transgene. Genomic rearrangements or deletions of parts of the host genome were also detected. Interestingly, the team was able to identify clones that originated from the same mother clone as they contained the same single integration region on a specific chromosome and had identical TLA coverage profiles. Stadermann et al. also observed a relationship between the product titer/process behavior and the (heavy and light chain) copy number in cells originating from the same mother clone. Clones with the same transgene integration site also had similar gene copy numbers of the antibody heavy and light chain and exhibited similar process performance. On the other hand, the clones with different genomic loci exhibited unrelated copy numbers and product titer.
In the clones with lower expression of heavy and light chains compared to other “top-producers” but optimal in process and platform fit, the team investigated the effect of supertransfection on process performance. Single-cell cloning of the supertransfected cell pools was performed to generate four monoclonal cell lines. The team observed that the resulting cell lines exhibited an up to 35% increase in antibody production, despite a lower process performance. TLA data analysis revealed that the resulting clones exhibited the same transgene integration site (with identical genomic coordinates and coverage patterns) as the original clone. The team also observed that the clones with boosted performance due to supertransfection exhibited a second integration site different from that of the original site of the mother clone. In addition, varying copy numbers of the expression constructs and genomic rearrangements were detected.
Stadermann et al. were curious as to which protein-coding genes were truncated, deleted, or translocated during the integration process as this could alter gene expression and CHO cell function. The team decided to investigate the impact of transgene integration on the CHO cell genome by annotating and classifying the different genes spanning the integration site or potentially deleted regions. Functional classification revealed that affected genes were mostly involved in cell division and growth such as the potent oncogene, Ei4e3 (Eukaryotic translation initiation factor 4E type 3), Diap2 (Mouse protein diaphanous homolog 2), and Rerg (Ras-related and estrogen-regulated growth inhibitor). These genes may have contributed to the improved process fit of the clones and superior phenotypes.
Overall, this study demonstrates the value of using TLA to characterize CHO production clones generated by random integration of expression cassettes during cell line development. Early in the process, this technique can support the selection of which CHO cells should be progressed based on the transgene integration sites and the genetic fingerprints, besides the phenotypic evaluation. By leveraging Genedata Selector® software to convert TLA data into actionable scientific insights, biopharma companies can benefit from better-informed go/no-go decisions and accelerate the screening of production cell lines. Accurate monitoring and confirmation of the integrity and stability of production clones allow for the successful and cost-effective development of effective therapies. To learn more about this study, read the publication.