A new strategy for dynamic metabolic flux estimation by integrating transient metabolome data into genome-scale metabolic models.

Abstract

Metabolic flux analysis (MFA) is a powerful tool for studying microbial cell physiology. Isotope-based MFA is the accepted standard for studying metabolic fluxes under steady-state conditions. However, its application under dynamic extracellular conditions is limited due to lack of proper techniques, such as rapid sampling and quenching, high cost and laborious execution. Here, we propose a new strategy to tackle this through incorporating dynamic metabolite abundance data into genome-scale metabolic models (GEM). First, a dummy extracellular pool concept is proposed for each dynamically changing metabolite, which represents a "sink" or "source", with corresponding dummy reactions coded into the GEM model. The dynamic model (expressed as differential equations) is then transformed into a quasi-steady-state model (expressed as linear equations), which can be easily solved by constraining the GEM model with the dynamic metabolite quantification data. For this, common linear-programming optimization algorithms were utilized to estimate the dynamic fluxes. Dynamic high-accuracy metabolite abundance data were obtained through the Isotope Dilution Mass Spectrometry (IDMS) method and high-speed sampling-quenching, and it was demonstrated that the newly proposed strategy could be successfully applied to obtain intracellular dynamic fluxes of Aspergillus niger under regimes of single and periodic extracellular glucose pulses. The applicability of the new method was also tested on dynamic fluxes estimation in a glucose pulse-response study of Saccharomyces cerevisiae. The proposed method provides a powerful tool to investigate cell physiology under dynamic conditions, especially relevant for bioprocess scale-up to industrial-scale bioreactors.