DI-UMONS : Dépôt institutionnel de l’université de Mons

Recherche transversale
Rechercher
(titres de publication, de périodique et noms de colloque inclus)
2013-06-06 - Colloque/Présentation - poster - Anglais - 1 page(s)

Didion David , Hantson Anne-Lise , "Optimizing the productivity of a recombinant protein producing CHO cell line by developing real-time analytical techniques for monitoring bioprocess" in NIR2013, La Grande Motte, France, 2013

  • Codes CREF : Biotechnologie (DI3800), Fermentation biosynthèse (DI3810), Chimie analytique (DI1314)
  • Unités de recherche UMONS : Chimie et Biochimie appliquées (F504)
  • Instituts UMONS : Institut des Biosciences (Biosciences)
  • Centres UMONS : Biosys (BIOSYS)
Texte intégral :

Abstract(s) :

(Anglais) Introduction: Biopharmaceutical products are industrially produced by genetically modified cells grown in bioreactors. Optimizing the productivity of a recombinant CHO cell line is the subject of a PhD thesis in progress in the Applied Chemistry and Biochemistry department of the University of Mons. This optimization is intended by monitoring the bioprocess with NIR spectroscopy and implementing a control loop to adapt in real-time the medium composition through the injection of key nutrients. Results and discussion: So far predictive models have been successfully developed with off-line collected spectra of synthetic solutions of main nutriments (glucose and glutamine) and inhibitors (lactate and ammonium ions) dissolved in cell culture medium. Spectra have been obtained using a Bruker™ Matrix-F FT-NIR spectrophotometer and a Hellma™ Immersion transflectance probe. The PLS toolbox designed by Eigenvector™ has been used for calibration and validation. When preparing the synthetic solutions, care has been taken to reject collinearity between analytes concentrations and to cover homogeneously the concentration ranges observed through the bioprocess. The intended real-time and segmented supplementation of the cell culture medium requires the analytical technique to enable the detection of concentration levels of nutriments limiting the protein synthesis. The LOQ has been evaluated for each analyte and for each set of measurement conditions as it is influenced by the component to be quantified and the presence of other components due to peaks overlapping and analyte-nonspecific spectral features. Models have been developed with 49 solutions of glucose and lactate on the one hand, and 49 solutions of glutamine and ammonium ions on the other hand. For glucose and lactate, the RMSECV and LOQ are around 0.05 g/l and 0.12 g/l respectively whereas the RMSECV and LOQ are around 0.2 mM and 0.6 mM respectively for glutamine and ammonium. Subsequently, the effects of increasing the number of analytes present in varying concentrations, adding varying amounts of cells and manipulating the optic fiber cables have been assessed independently. RMSECV and LOQ related to models developed with 100 solutions of the four analytes have roughly doubled in comparison with previous models. The effect of manipulating the optic fibers has been resolved by developing models with spectra of various origins whereas the influence of the cells on the predictive abilities of the models has appeared to be minor. Conclusion: The effort of calibration of the NIRS has confirmed the possibility of monitoring four analytes in synthetic solutions containing mammalian cells and matrix effects associated to the cell culture medium. Encouragingly the LOQ of the analytes have been estimated inferior to the limiting concentrations levels reported in the literature. However the solutions that have been measured so far are by far less complex than the bioprocess intended to be monitored in fine. The model designed to monitor the bioprocess will be developed with calibration spectra collected in solutions with varying concentrations of both the desired analytes and all potentially interfering compounds. Novelty statement: The NIRS monitoring is planned to be extended to the amino acids limiting or inhibiting the protein production. The real-time and in situ monitoring progresses will allow the design of a dynamic feeding strategy that meets the real needs of the cells grown in the bioreactor. Summary: The numerous assets of the NIRS (real-time, in situ and multi-analyte quantification) are intended to benefit the optimization of the productivity of CHO cells. The analytes that limit the protein production must be identified and included in the calibration model enabling us to monitor the bioprocess. With that purpose in mind, models of increasing complexity have been developed in order to assess the ability of the NIRS to monitor components present at low concentrations, below 1 mM. On basis of these developments, a control strategy of the cell culture medium composition will be designed.