High-throughput FTIR-based bioprocess analysis of recombinant cyprosin production

• Published in: Journal of industrial microbiology & biotechnology 2017, Vol.44 (1), p.49-61
• Main Authors: Sampaio, Pedro N., Sales, Kevin C., Rosa, Filipa O., Lopes, Marta B., Calado, Cecília R. C.
• Format: Article
• Language: English
• Subjects: Aspartic Acid Endopeptidases - biosynthesis; Biochemistry; Bioinformatics; Biomass; Biomedical and Life Sciences; Biosynthesis; Biotechnology; Biotechnology - methods; Cell culture; Cell Culture and Bioengineering - Original Paper; Data analysis; Ethanol; Ethanol - metabolism; Fermentation; Galactose; Genetic Engineering; Genetic recombination; Glucose; Glucose - analysis; Inorganic Chemistry; Knowledge; Least-Squares Analysis; Life Sciences; Mean square errors; Microbiology; Optimization; Principal Component Analysis; Principal components analysis; Recombinant Proteins - biosynthesis; Regression Analysis; Saccharomyces cerevisiae; Spectrophotometry, Infrared; Spectroscopy; Spectroscopy, Fourier Transform Infrared - methods; Spectrum analysis; Studies; Temperature; Trends; Yeast; Yeasts
• ISSN: 1367-5435; 1476-5535
• DOI: 10.1007/s10295-016-1865-0

To increase the knowledge of the recombinant cyprosin production process in Saccharomyces cerevisiae cultures, it is relevant to implement efficient bioprocess monitoring techniques. The present work focuses on the implementation of a mid-infrared (MIR) spectroscopy-based tool for monitoring the recombinant culture in a rapid, economic, and high-throughput (using a microplate system) mode. Multivariate data analysis on the MIR spectra of culture samples was conducted. Principal component analysis (PCA) enabled capturing the general metabolic status of the yeast cells, as replicated samples appear grouped together in the score plot and groups of culture samples according to the main growth phase can be clearly distinguished. The PCA-loading vectors also revealed spectral regions, and the corresponding chemical functional groups and biomolecules that mostly contributed for the cell biomolecular fingerprint associated with the culture growth phase. These data were corroborated by the analysis of the samples’ second derivative spectra. Partial least square (PLS) regression models built based on the MIR spectra showed high predictive ability for estimating the bioprocess critical variables: biomass ( R 2  = 0.99, RMSEP 2.8%); cyprosin activity ( R 2  = 0.98, RMSEP 3.9%); glucose ( R 2  = 0.93, RMSECV 7.2%); galactose ( R 2  = 0.97, RMSEP 4.6%); ethanol ( R 2  = 0.97, RMSEP 5.3%); and acetate ( R 2  = 0.95, RMSEP 7.0%). In conclusion, high-throughput MIR spectroscopy and multivariate data analysis were effective in identifying the main growth phases and specific cyprosin production phases along the yeast culture as well as in quantifying the critical variables of the process. This knowledge will promote future process optimization and control the recombinant cyprosin bioprocess according to Quality by Design framework.