Résumé | More than half of licensed therapeutic recombinant proteins (r‐proteins) are manufac-tured using constitutively‐expressing, stably‐transfected Chinese hamster ovary (CHO)clones. While constitutive CHO expression systems have proven their efficacy for themanufacturing of monoclonal antibodies, many next‐generation therapeutics such ascytokines and bispecific antibodies as well as biological targets such as ectodomains oftransmembrane receptors remain intrinsically challenging to produce. Herein, weexploited a cumate‐inducible CHO platform allowing reduced expression of variousclasses of r‐proteins during selection of stable pools. Following stable pool generation,fed‐batch productions showed that pools generated without cumate (OFF‐pools) weresignificantly more productive than pools selected in the presence of cumate (ON‐pools)for 8 out of the 10 r‐proteins tested, including cytokines, G‐protein coupled receptors(GPCRs), the HVEM membrane receptor ectodomain, the multifunctional protein HighMobility Group protein B1 (HMGB1), as well as monoclonal and bispecific T‐cell engagerantibodies. We showed that OFF‐pools contain a significantly larger proportion of cellsproducing high levels of r‐proteins and that these cells tend to proliferate faster whenexpression is turned off, suggesting that r‐protein overexpression imposes a metabolicburden on the cells. Cell viability was lower and pool recovery was delayed duringselection of ON‐pools (mimicking constitutive expression), suggesting that high producerswere likely lost or overgrown by faster‐growing, low‐producing cells. We also observed acorrelation between the expression levels of the GPCRs with Binding immunoglobulinProtein, an endoplasmic reticulum (ER) stress marker. Taken together, these data suggestthat using an inducible system to minimize r‐protein expression during stable CHO poolselection reduces cellular stresses, including ER stress and metabolic burden, leading topools with greater frequency of high‐expressing cells, resulting in improved volumetricproductivity. |
---|