Intracellular product recycling in high succinic acid producing yeast at low pH

Abstract

Background: The metabolic engineering of Saccharomyces cerevisiae for the production of succinic acid has progressed dramatically, and a series of high-producing hosts are available. At low cultivation pH and high titers, the product transport can become bidirectional, i.e. the acid is reentering the cell and is again exported or even catabolized. Here, a quantitative approach for the identification of product recycling fluxes is developed.

Results: The metabolic flux distributions at two time-points of the fermentation process were analyzed. ¹³C labeled succinic acid was added to the extracellular space and intracellular enrichments were measured and subsequently used for the estimation of metabolic fluxes. The labeling was introduced by a labeling switch experiment, leading to an immediate labeling of about 85% of the acid while keeping the total acid concentration constant. Within 100 s significant labeling enrichment of the TCA cycle intermediates fumarate, iso-citrate and α-ketoglutarate was observed, while no labeling was detected for malate and citrate. These findings suggest that succinic acid is rapidly exchanged over the cellular membrane and enters the oxidative TCA cycle. Remarkably, in the oxidative direction malate ¹³C enrichment was not detected, indicating that there is no flux going through this metabolite pool. Using flux modeling and thermodynamic assumptions on compartmentation it was concluded that malate must be predominantly cytosolic while fumarate and iso-citrate were more dominant in the mitochondria.

Conclusions: Adding labeled product without changing the extracellular environment allowed to quantify intracellular metabolic fluxes under high producing conditions and identify product degradation cycles. In the specific case of succinic acid production, compartmentation was found to play a major role, i.e. the presence of metabolic activity in two different cellular compartments lead to intracellular product degradation reducing the yield. We also observed that the flux from glucose to succinic acid branches at two points in metabolism: (1) At the level of pyruvate, and (2) at cytosolic malate which was not expected.

Keywords: 13C labeling; Metabolic flux analysis; Saccharomyces cerevisiae; Succinic acid production.

Fig. 1

Setup of the bioscope labeling experiment. a The bioscope has inflows from the bioreactor (flow 0.14 mL/min) and the BioScope feed (0.56 mL/min). b The BioScope feed contains labeled succinic acid matching the concentration of succinic acid of the bioreactor broth. The solution also matches the residual glucose concentration (unlabeled). This setup leads to a dilution of the biomass concentration while the residual concentration of glucose and succinic acid are similar to the BR cultivation. Using labeled succinic acid allows for immediate labeling with about 80% enrichment

Fig. 2

Estimated biomass specific rates over the cultivation time. Table 2 displays the rates at the time of the bioscope experiments. Standard deviations are calculated using linear error propagation (i.e. linearization of the ODE model)

Fig. 3

Scheme of intracellular transport reactions of TCA cycle intermediates. A complete list of assumed transport reactions can be found in Table 1, estimated rates are displayed in Figs. 5 and 7

Fig. 4

Enrichment measurements and mass isotopomer simulation for t = 52 h. The mass isotopomer simulation represents the sum of the mitochondrial and cytosolic prediction. The measurements for Oaa were less accurate because of the low intracellular concentration and were included in the parameter estimation with a 10× higher standard deviation. The mass isotopomer measurements can be found in Additional file 1: Table S6

Fig. 5

Metabolic flux distribution at t = 52 h. The metabolic fluxes were identified by the addition of labeled succinic acid and parameter estimation. Please note that not all reactions are shown to facilitate the overview. A complete list of reactions can be found in the Additional file 1: Tables S1, S2

Fig. 6

Simulated and measured mass isotopomers for t = 73 h. From the simulation the sum of mass isotopomers from mitochondria and cytosol are calculated, representing the predicted whole-cell measurements. The mass isotopomer measurements can be found in Additional file 1: Table S7

Fig. 7

Estimated flux distribution at t = 73 h. All flux values are given in µmol/g/h. Please note that not all reactions are shown to facilitate the overview. A complete list of reactions can be found in the Additional file 1: Tables S1, S3