Recombinant IgG expression in mammalian cells

Abstract

IgG are heterotetrameric proteins with two identical heavy (H) and two identical light (L) chain, which are now one of the largest categories of biopharmaceutical products in development. Mammalian cells are the dominant system for the production of recombinant proteins for clinical applications. The volumetric productivity of mammalian cells cultivated in bioreactors increased significantly over the past 20 years because of improvements in media composition and process control. However, there are only few ideas to predict what kind of vector design is optimal for stable or transient production of recombinant IgG. In this study, a series of mammalian expression vectors for the production of recombinant human or chimeric IgG antibodies with different expression cassette designs were constructed. The impact of promoters, the effects of the ratio of heavy chain to light chain for efficient IgG folding and assembly, the difference between monocistronic and polycistronic mammalian expression as well as the impact of bicistronic expression vector on the stability of long term recombinant IgG expression were systematically analyzed. The impact of cistron arrangement of H and L chain in polycistronic vectors was firstly evaluated in transfected cells using yellow fluorescence protein as reporter. No apparent affection was detected from the cistron arrangements of H and L chain in transiently transfected HEK 293T cells. However, in stably transfected CHO-K1 cells, the downstream cistron of H chain exerted more negative affection to cap-dependent translation of upstream cistron than L chain did. Promoters effect IgG expression levels significantly in transiently transfected cells, the EF1-alpha promoter was more efficient than the human CMV-IE promoter for recombinant IgG expression. The Encephalomyocarditis virus (EMCV) IRES mediated bicistronic expression constructs could yield IgG expression levels in the same range as monocistronic expression in transiently transfected cells, and it could keep long-term stability in CHO-K1 cell lines. To investigate the optimal range of the H and L chain intracellular expression levels for IgG folding and assembly, a series of EMCV IRES variants with different translation efficiencies was used to mediate H chain translation in the bicistronic setup L-IRES-H with IgG L chain as first cistron. It provides a novel method for the investigation of oligomeric protein mammalian expression. It seems that the optimal expression level was obtained when the IRES mediated translation efficiency was about 0,5 compared to cap-dependent translation. To enhance the stability of IgG expression in stably transfected CHO cells, a 2,2 Kb scaffold/matrix attached region (S/MAR) from human beta-interferon gene was used to modify the bicistronic IgG expression plasmid. It was shown that the S/MAR could apparently increase the prevalence of positive clones when it was placed flanking the bicistronic expression cassette. The effect on the expression levels from single clones needs to be studied further. Finally, a human anti-MUC1 scFv was transformed into IgG and produced in CHO cells. Expression levels were evaluated in transiently and stably transfected cells.