Singh G.P.,Massachusetts Institute of Technology |
Goh S.,Bioprocessing Technology Institute BTI |
Canzoneri M.,Sanofi S.A. |
Ram R.J.,Massachusetts Institute of Technology
Journal of Raman Spectroscopy | Year: 2015
We demonstrate the detection of glucose and lactate concentrations with high accuracy in the supernatants of Chinese hamster ovary (CHO) cell culture, grown in shake flasks in batch fermentation mode, using Raman spectroscopy and explicit model-based classical least squares (CLS) algorithm. A deterministic Raman spectral library of pure components was created by acquiring Raman spectra from credible nutrient media constituents and CHO cell culture metabolites. Only analytes present with concentration above the instrument detection limit were included in this library. Residuals obtained after CLS analyses were used to identify missing components and to generate a revised library. An algorithmic sieve was thus construed to obtain an appropriate Raman spectral library from a complex chemical mixture that is well-defined but an industrial secret. High performance liquid chromatography (HPLC) was used to provide reference glucose and lactate concentrations. [Correction added on 30 April 2015, after first online publication: the last three sentences of the abstract (previously stated in the abstract) have been deleted]. Copyright © 2015 John Wiley & Sons, Ltd.