Acetate Metabolism and the Inhibition of Bacterial Growth by Acetate

Abstract

During aerobic growth on glucose, Escherichia coli excretes acetate, a mechanism called "overflow metabolism." At high concentrations, the secreted acetate inhibits growth. Several mechanisms have been proposed for explaining this phenomenon, but a thorough analysis is hampered by the diversity of experimental conditions and strains used in these studies. Here, we describe the construction of a set of isogenic strains that remove different parts of the metabolic network involved in acetate metabolism. Analysis of these strains reveals that (i) high concentrations of acetate in the medium inhibit growth without significantly perturbing central metabolism; (ii) growth inhibition persists even when acetate assimilation is completely blocked; and (iii) regulatory interactions mediated by acetyl-phosphate play a small but significant role in growth inhibition by acetate. The major contribution to growth inhibition by acetate may originate in systemic effects like the uncoupling effect of organic acids or the perturbation of the anion composition of the cell, as previously proposed. Our data suggest, however, that under the conditions considered here, the uncoupling effect plays only a limited role.IMPORTANCE High concentrations of organic acids such as acetate inhibit growth of Escherichia coli and other bacteria. This phenomenon is of interest for understanding bacterial physiology but is also of practical relevance. Growth inhibition by organic acids underlies food preservation and causes problems during high-density fermentation in biotechnology. What causes this phenomenon? Classical explanations invoke the uncoupling effect of acetate and the establishment of an anion imbalance. Here, we propose and investigate an alternative hypothesis: the perturbation of acetate metabolism due to the inflow of excess acetate. We find that this perturbation accounts for 20% of the growth-inhibitory effect through a modification of the acetyl phosphate concentration. Moreover, we argue that our observations are not expected based on uncoupling alone.

Keywords: Escherichia coli; acetate; acetate metabolism; growth inhibition; metabolic flux analysis; overflow metabolism.

FIG 1

Schematic representation of the major metabolic pathways of acetate metabolism. Acetate can be generated directly from pyruvate by decarboxylation (using the enzyme pyruvate oxidase, PoxB) or from acetyl-CoA via the intermediate acetyl-phosphate (Ac∼P) (reactions catalyzed by the enzymes phosphotransacetylase, Pta, and acetate kinase, AckA). Acetate can freely diffuse across the cell membrane in the protonated form (13), HAc, or as an acetate ion, Ac⁻, via the transporter ActP or SatP (38). Protons that enter the cell in the form of HAc can be expelled at the expense of energy. The enzyme acetyl-CoA synthetase, Acs, efficiently converts intracellular acetate into acetyl-CoA. Acetate is also involved in several metabolic pathways. For example, the biosynthesis of methionine is inhibited by acetate. Excess intracellular acetate perturbs the anion balance in the cell and thus could inhibit other metabolic reactions (20). The pool of the major intracellular anion, glutamate, is strongly reduced when the intracellular concentration of acetate is high.

FIG 2

Growth inhibition by acetate. The inoculated culture is split and cultured in separate shake flasks containing minimal medium with glucose at pH 7.4. A small volume of minimal medium with acetate is added to one culture, and an identical volume of minimal medium without acetate is added to the second, control culture. The moment of acetate addition is indicated by the arrow. The optical density measurements represent the means from three experiments. The error bars (mostly smaller than, and therefore hidden by, the circles showing the data points) are two times the standard errors of the means. The data shown correspond to the wild-type strain and a final concentration of acetate of 128 mM.

FIG 3

Growth inhibition by different concentrations of acetate and different pH levels. Experiments like those shown in Fig. 2 were carried out for different concentrations of acetate at pH 7.4 (blue dots) and pH 6.4 (red dots). Each data point shows the mean growth rate from three or four independent experiments, as well as an error bar equal to twice the standard errors of the means. An exponential function with baseline of 0.3 h⁻¹ was fit to the data (blue and red curves). The baseline approaches the growth rate in minimal medium with 128 mM acetate as the sole carbon source (green line). For reference, the growth rate on 3 g liter⁻¹ of acetate, corresponding to 50 mM, is also shown (green dot). Notice that for some measurements the error bar is so small that it coincides with the measurement dot.

FIG 4

Growth rate of mutant strains in the absence of acetate and in the presence of 128 mM acetate. The growth rate of all mutant strains was measured in a standard shake flask culture as described for Fig. 2 and computed from the data as described in Materials and Methods. We report the means from at least three independent experiments. The error bars represent twice the standard errors of the means. The growth rates without and with acetate are shown in panels a and b, respectively, and the inhibition index is in panel c. The asterisks (* and **) in panels b and c indicate growth rates and inhibition indices for mutant strains that are significantly different from the growth rate and inhibition index of the wild-type strain, for significance thresholds of 0.06 and 0.01, respectively (see Materials and Methods).

FIG 5

Measurement of metabolites taken up and excreted by E. coli. Bacteria were grown in a shake flask, as described for Fig. 2, and samples were removed at regular time intervals and analyzed for the different metabolites (see Materials and Methods). The left column shows cultures grown on glucose alone. Acetate (128 mM) was added to the cultures in the right column after 3 h of growth (indicated by the vertical dashed line). The metabolite measured is indicated on top of each row. The measurements were carried out in quadruplicate in the wild-type strain (green), the Δacs pta mutant (red), and the Δacs pta ackA mutant (blue). The error bars indicate twice the standard errors of the means. For reference, OD₆₀₀ curves are shown as dashed lines. Note that the ordinate scale is smaller for the metabolites in the bottom four rows, and the ordinate scale of acetate measurements with added acetate is much larger than the others.

FIG 6

Changes in metabolic fluxes in central carbon metabolism predicted by a genome-scale model of E. coli. The measurements of extracellular metabolites shown in Fig. 5 were used to compute uptake and secretion rates under each of the conditions, as outlined in Materials and Methods. These rates were used to constrain the exchange fluxes of a genome-scale model of E. coli metabolism (29), while the rate of the biomass reaction was set to the experimentally determined growth rate. We used metabolic flux analysis to define a space of solutions consistent with the measured fluxes and the stoichiometry structure of the metabolic network. This solution space was sampled in a random and unbiased manner using a Monte Carlo approach (see Materials and Methods). (a) Scatter plot of the predicted fluxes for 64 reactions in central carbon metabolism of the wild-type E. coli strain growing on glucose, in the absence and presence of acetate. Under each condition, the fluxes have been normalized by the growth rate. The scatter plot shows a very strong correlation between the predicted flux distributions in the absence and presence of acetate, with all reactions clustered around the diagonal (R² = 0.89). (b) Same as panel a, except the Δacs pta strain (R² = 0.94) was examined. (c) Same as panel a, except the Δacs pta ackA strain (R² = 0.94) was examined.

FIG 7

Biomass yield of wild-type mutant strains in the absence of acetate and in the presence of 128 mM acetate. The biomass yields of the wild-type strain (green), the Δacs pta mutant (red), and the Δacs pta ackA mutant (blue), in the absence of acetate and in the presence of 128 mM acetate, were measured as described in Materials and Methods. We report the means from four independent experiments. The error bars represent twice the standard errors of the means. The diagonal corresponds to identical biomass yields in the presence and absence of acetate. Although the measurement means are located below the diagonal, suggesting a lower biomass yield in the presence of acetate, the pairwise differences in yield with and without acetate are not statistically significant at confidence levels of 0.01 and 0.06 (see Materials and Methods).