Molecular portrait of high productivity in recombinant NS0 cells

• Published in: Biotechnology and bioengineering 2007-07, Vol.97 (4), p.933-951
• Main Authors: Seth, Gargi, Philp, Robin J, Lau, Ally, Jiun, Kok Yee, Yap, Miranda, Hu, Wei-Shou
• Format: Article
• Language: English
• Subjects: Animals; Biological and medical sciences; Biotechnology; Cell Line, Tumor; Cellular biology; Computational Biology; Deoxyribonucleic acid; DNA; Electrophoresis, Gel, Two-Dimensional; Fundamental and applied biological sciences. Psychology; Gene expression; Gene Expression Regulation, Neoplastic; Gene Regulatory Networks; Genetic recombination; genomics; iTRAQ; mammalian cell culture; Mice; microarray; mouse myeloma; Multiple Myeloma - genetics; Multiple Myeloma - metabolism; Myeloma Proteins - genetics; Myeloma Proteins - metabolism; Neoplasm Proteins - genetics; Neoplasm Proteins - metabolism; NS0 cells; Oligonucleotide Array Sequence Analysis; Online Systems; Proteins; Proteome - analysis; Proteomics; recombinant protein production; Recombinant Proteins - biosynthesis; Sensitivity and Specificity; Studies; Tandem Mass Spectrometry; Transcription, Genetic; transcriptome; two dimensional gel electrophoresis
• ISSN: 0006-3592; 1097-0290
• DOI: 10.1002/bit.21234

Many important therapeutic proteins are produced in recombinant mammalian cells. Upon the introduction of the product gene, the isolated clones typically exhibit a wide range of productivity and high producers are subsequently selected for use in production. Using DNA microarray, two-dimensional gel electrophoresis (2DE), and iTRAQ™ as global surveying tools, we examined the transcriptome and proteome profiles of 11 lines of NS0 cells producing the same antibody molecule. Genes that are significantly differentially expressed between high and low producer groups statistically fall into a number of functional classes. Their distribution among the functional classes differs somewhat between transcriptomic and proteomic results. Overall, a high degree of consistency between transcriptome and proteome analysis are seen, although some genes exhibiting inconsistent trends between transcript and protein levels were observed as expected. In a novel approach, functional gene networks were retrieved using computational pathway analysis tools and their association with productivity was tested by physiological comprehension of the possible pathways involved in high recombinant protein production. Network analysis indicates that protein synthesis pathways were altered in high producers at both transcriptome and proteome levels, whereas the effect on cell growth/death pathways was more prominent only at the transcript level. The results suggest a common mechanism entailing the alteration of protein synthesis and cell growth control networks leading to high productivity. However, alternate routes with different sets of genes may be invoked to give rise to the same mechanistic outcomes. Such systematic approaches, combining transcriptomic and proteomic tools to examine high and low producers of recombinant mammalian cells will greatly enhance our capability to rationally design high producer cells. This work is a first step towards shedding a new light on the global physiological landscape of hyper productivity of recombinant cells. Biotechnol. Bioeng. 2007;97: 933-951.