Metabolic adjustment upon repetitive substrate perturbations using dynamic 13C-tracing in yeast

Abstract

Background: Natural and industrial environments are dynamic with respect to substrate availability and other conditions like temperature and pH. Especially, metabolism is strongly affected by changes in the extracellular space. Here we study the dynamic flux of central carbon metabolism and storage carbohydrate metabolism under dynamic feast/famine conditions in Saccharomyces cerevisiae.

Results: The metabolic flux reacts fast and sensitive to cyclic perturbations in substrate availability. Compared to well-documented stimulus-response experiments using substrate pulses, different metabolic responses are observed. Especially, cells experiencing cyclic perturbations do not show a drop in ATP with the addition of glucose, but an immediate increase in energy charge. Although a high glycolytic flux of up to 5.4 mmol g _DW^-1 h^-1 is observed, no overflow metabolites are detected. From famine to feast the glucose uptake rate increased from 170 to 4788 μmol g _DW^-1 h^-1 in 24 s. Intracellularly, even more drastic changes were observed. Especially, the T6P synthesis rate increased more than 100-fold upon glucose addition. This response indicates that the storage metabolism is very sensitive to changes in glycolytic flux and counterbalances these rapid changes by diverting flux into large pools to prevent substrate accelerated death and potentially refill the central metabolism when substrates become scarce. Using ¹³C-tracer we found a dilution in the labeling of extracellular glucose, G6P, T6P and other metabolites, indicating an influx of unlabeled carbon. It is shown that glycogen and trehalose degradation via different routes could explain these observations. Based on the ¹³C labeling in average 15% of the carbon inflow is recycled via trehalose and glycogen. This average fraction is comparable to the steady-state turnover, but changes significantly during the cycle, indicating the relevance for dynamic regulation of the metabolic flux.

Conclusions: Comparable to electric energy grids, metabolism seems to use storage units to buffer peaks and keep reserves to maintain a robust function. During the applied fast feast/famine conditions about 15% of the metabolized carbon were recycled in storage metabolism. Additionally, the resources were distributed different to steady-state conditions. Most remarkably is a fivefold increased flux towards PPP that generated a reversed flux of transaldolase and the F6P-producing transketolase reactions. Combined with slight changes in the biomass composition, the yield decrease of 5% can be explained.

Keywords: 13C labeling; Dynamic fluxes; Metabolomics; Multiple perturbations; Storage carbohydrates; Systems biology; Yeast.

Fig. 1

Profile of the experimental feeding and sampling regime. After a chemostat phase (reference steady-state), a block-wise feed is applied at the same average substrate supply and dilution rate. Intracellular concentrations are measured by sampling two cycles, then the feed is switched to labeled substrate and the enrichment is monitored for three consecutive cycles

Fig. 2

Measured and simulated concentration and ¹³C-labeling enrichment (C-molar average) of glycolytic metabolites during three consecutive feast/famine cycles

Fig. 3

Measured and simulated concentration and ¹³C-labeling enrichment (C-molar average) of PPP metabolites during three consecutive feast/famine cycles

Fig. 4

Measured and simulated concentration and ¹³C-labeling enrichment (C-molar average) of metabolites of the storage carbohydrate branches during three consecutive feast/famine cycles. Glycogen and extracellular trehalose ¹³C enrichment was unfortunately not measured

Fig. 5

Estimated dynamic fluxes (Glycolysis) during the feast/famine regime. Horizontal lines represent: average flux over the cycle (blue) and steady-state flux a D = 0.1 h⁻¹ (black)

Fig. 6

Dynamic fluxes of the pentose phosphate pathway during the feast/famine conditions. Horizontal lines represent: average flux over the cycle (black) and steady-state flux a D = 0.1 h⁻¹ (blue)

Fig. 7

Estimated dynamic flux of storage synthesis and degradation pathways during the feast/famine regime. Horizontal lines represent: average flux over the cycle (black) and steady-state flux a D = 0.1 h⁻¹ (blue)

Fig. 8

Action mass ratio (Q) for the PGI reaction. Two different values for Keq reported in the literature are shown: dashed line [14] and dotted line [30]

Fig. 9

Comparison of flux distribution, on the left the average over the feast/famine cycle, on the right the reference steady-state (adapted from [32]). All values are given in µmol gDW-1 h⁻¹