Maximizing protein production by keeping cells at optimal secretory stress levels using real-time control approaches.
Sebastián Sosa-CarrilloHenri GalezSara NapolitanoFrançois BertauxGregory BattPublished in: Nature communications (2023)
Optimizing the production of recombinant proteins is a problem of major industrial and pharmaceutical importance. Secretion of the protein by the host cell considerably simplifies downstream purification processes. However, for many proteins, this is also the limiting production step. Current solutions involve extensive engineering of the chassis cell to facilitate protein trafficking and limit protein degradation triggered by excessive secretion-associated stress. Here, we propose instead a regulation-based strategy in which induction is dynamically adjusted to an optimal strength based on the current stress level of the cells. Using a small collection of hard-to-secrete proteins, a bioreactor-based platform with automated cytometry measurements, and a systematic assay to quantify secreted protein levels, we demonstrate that the secretion sweet spot is indicated by the appearance of a subpopulation of cells that accumulate high amounts of proteins, decrease growth, and face significant stress, that is, experience a secretion burnout. In these cells, adaptations capabilities are overwhelmed by a too strong production. Using these notions, we show for a single-chain antibody variable fragment that secretion levels can be improved by 70% by dynamically keeping the cell population at optimal stress levels using real-time closed-loop control.
Keyphrases
- induced apoptosis
- single cell
- cell cycle arrest
- protein protein
- amino acid
- cell therapy
- binding protein
- machine learning
- wastewater treatment
- high throughput
- signaling pathway
- body mass index
- heavy metals
- physical activity
- deep learning
- mesenchymal stem cells
- risk assessment
- weight loss
- pi k akt
- oxidative stress
- bone marrow