Integrated cell and process engineering for improved transient production of a "difficult-to-express" fusion protein by CHO cells

Based on an optimized electroporation protocol, we designed a rapid, milliliter-scale diagnostic transient production assay to identify limitations in the ability of Chinese hamster ovary (CHO) cells to produce a model difficult-to-express homodimeric Fc-fusion protein, Sp35Fc, that exhibited very low volumetric titer and intracellular formation of disulfide-bonded oligomeric aggregates post-transfection. As expression of Sp35Fc induced an unfolded protein response in transfected host cells, we utilized the transient assay to compare, in parallel, multiple functionally diverse strategies to engineer intracellular processing of Sp35Fc in order to increase production and reduce aggregation as two discrete design objectives. Specifically, we compared the effect of (i) co-expression of ER-resident molecular chaperones (BiP, PDI, CypB) or active forms of UPR transactivators (ATF6c, XBP1s) at varying recombinant gene load, (ii) addition of small molecules known to act as chemical chaperones (PBA, DMSO, glycerol, betaine, TMAO) or modulate UPR signaling (PERK inhibitor GSK2606414) at varying concentration, (iii) a reduction in culture temperature to 32 degrees C. Using this information, we designed a biphasic, Sp35Fc-specific transient manufacturing process mediated by lipofection that utilized CypB co-expression at an optimal Sp35Fc:CypB gene ratio of 5:1 to initially maximize transfected cell proliferation, followed by addition of a combination of PBA (0.5mM) and glycerol (1% v/v) at the onset of stationary phase to maximize cell specific production and eliminate Sp35Fc aggregation. Using this optimal, engineered process transient Sp35Fc production was significantly increased sixfold over a 12 day production process with no evidence of disulfide-bonded aggregates. Finally, transient production in clonally derived sub-populations (derived from parental CHO host) screened for a heritably improved capability to produce Sp35Fc was also significantly improved by the optimized process, showing that protein-specific cell/process engineering can provide a solution that exceeds the limits of genetic/functional diversity within heterogeneous host cell populations. Biotechnol. Bioeng. 2015;112: 2527-2542. (c) 2015 Wiley Periodicals, Inc.