Metabolic fingerprinting as a tool to monitor whole-cell biotransformations.

Fourier transform infrared (FT-IR) spectroscopy was employed as a rapid high-throughput phenotypic typing technique to generate metabolic fingerprints of Escherichia coli MG1655 pDTG601A growing in fed-batch culture, during the dioxygenase-catalysed biotransformation of toluene to toluene cis-glycol. With toluene fed as a vapour, the final toluene cis-glycol concentration was 83 mM, whereas the product concentration was only 22 mM when the culture was supplied with liquid toluene. Multivariate statistical analysis employing cluster analysis was used to analyse the dynamic changes in the data. The analysis revealed distinct trends and trajectories in cluster ordination space, illustrating phenotypic changes related to differences in the growth and product formation of the cultures. In addition, partial least squares regression was used to correlate the FT-IR metabolic fingerprints with the levels of toluene cis-glycol and acetate, the latter being an indicator of metabolic stress. We propose that this high-throughput metabolic fingerprinting approach is an ideal tool to assess temporal biochemical dynamics in complex biological processes, as demonstrated by this redox biotransformation. Moreover, this approach can also give useful information on product yields and fermentation health indicators directly from the fermentation broth without the need for lengthy chromatographic analysis of the products.