Classification and Characterization of CHO  Cell Lines on the Basis of Transcriptomic  Signatures

Chinese hamster ovary (CHO) cell lines were first used in 1919 in place of mice for typing pneumococci (Jayapal et al, 2007), and have later been transferred to research laboratories due to their ability to robustly grow, adapt to various culture conditions and ability to show resistance towards viral contamination. Also, CHO cell lines carry larger chromosomes (in comparison to human chromosome) that can easily be identified, tested and experimented with, therefore CHO cell lines quickly made their way to become a workhorse in many research laboratories for experimentation and testing. So far, a substantial amount of research has been done to understand media composition, culture studying cell cycle, toxicology and cytogenetic studies. Further, CHO cells are considered one of the robust recombinant DNA hosts for production of pharmaceutical proteins. Today, after almost six decades of research, CHO cell lines are still perceived as a robust system for the production of pharmaceutical therapeutic protein, due to their ability to perform apt posttranslational modification and produce product, compatible with human immune system. However, the, genetic basis of CHO cell lines to determine underlying mechanism behind their versatility has not been understood much in detail.

Total expression load of the genome is divided into various fragments to regulate various pathways. Pathways that receive maximum proportion of the expression load drive the functional efficiency of the cells and ultimately decide the fate of the cells.The scope of the present study lies in exploring detailed differences between expression load of various kinds of CHO quasispecies (e.g.CHO-S,CHO-DH44,CHO-K1)on the basis of expression data, derived from genome-wide transcriptome either using mRNA-seq or miRNA-seq. To determine the expression profiles classifying the distinction between some of the most desired pathways such as secretory, N-glycan, stress and lipid differentiation between producer and parental cell lines, were made at different levels. Further, special emphasis has been given to understand miRNA pathways by performing mRNA and miRNA correlation analysis.

In the first study, expression load on the gene sets driving the secretory machinery of the cells has been explored in detail and compared across various CHO cell lines to derive the differences among them. Further, distinction obtained in the form of pattern has been used to evaluate inherent characteristics of various CHO quasispecies. In addition, a public CHO RNA-seq database has been generated to illustrate genomic expression, obtained from transcriptomic data across various time points, of several CHO cell lines from the study.

In the second study, the ability to perform posttranslational modifications was explored in detail to discern the genomic expression patterns of N-glycan units as it is considered one of the most important trait of CHO cell lines. Further, an additional dimensionality has been supplemented to the study by adding expression information of miRNA targeting and later correlating it to the mRNA expression profile. Further study has allowed us to modify the N-glycan profile of the CHO cells to make them a more efficient host for producing therapeutic proteins. In conclusion, we found four miRNA with an effect on protein glycosylation and several highly expressed targets, obtained from producer and parental cell lines, as an asset to the CHO community for future miRNA engineering studies.
In the third study, the exploration of lipid and fatty acid-related pathways in addition to stress related pathways have been explored in industrial CHO cell lines producing human gastric lipase (hGL) to address the contribution of genomic expression and the driving force behind it. Pathways related to lipid hydrolysis and lipid biosynthesis were found to respond both when clones at various time points are analyzed separately and together, indicating, introduction of hGL impact the internal environment of the recombinant cell lines. This finding highlights that the interplay between mammalian proteins and mammalian expression hosts is not limited to the translation burden, but also the activity of the protein.

Overall, the work demonstrated in the thesis has given new dimension to visualize mRNA and miRNA based expression data to CHO community, in the form of targets and tools that can be utilized to understand and finally make CHO cell lines a better and improved workhorse.