Mechanisms of Acetoin Toxicity and Adaptive Responses in an Acetoin-Producing Species, Lactococcus lactis

Abstract

Acetoin, 3-hydroxyl,2-butanone, is extensively used as a flavor additive in food products. This volatile compound is produced by the dairy bacterium Lactococcus lactis when aerobic respiration is activated by haem addition, and comprises ∼70% of carbohydrate degradation products. Here we investigate the targets of acetoin toxicity, and determine how acetoin impacts L. lactis physiology and survival. Acetoin caused damage to DNA and proteins, which related to reactivity of its keto group. Acetoin stress was reflected in proteome profiles, which revealed changes in lipid metabolic proteins. Acetoin provoked marked changes in fatty acid composition, with massive accumulation of cycC19:0 cyclopropane fatty acid at the expense of its unsaturated C18:1 fatty acid precursor. Deletion of the cfa gene, encoding the cycC19:0 synthase, sensitized cells to acetoin stress. Acetoin-resistant transposon mutagenesis revealed a hot spot in the high affinity phosphate transporter operon pstABCDEF, which is known to increase resistance to multiple stresses. This work reveals the causes and consequences of acetoin stress on L. lactis, and may facilitate control of lactic acid bacteria production in technological processes. IMPORTANCE Acetoin, 3-hydroxyl,2-butanone, has diverse uses in chemical industry, agriculture, and dairy industries as a volatile compound that generates aromas. In bacteria, it can be produced in high amount by Lactococcus lactis when it grows under aerobic respiration. However, acetoin production can be toxic and detrimental for growth and/or survival. Our results showed that it damages DNA and proteins via its keto group. We also showed that acetoin modifies membrane fatty acid composition with the production of cyclopropane C19:0 fatty acid at the expense of an unsaturated C18:1. We isolated mutants more resistant to acetoin than the wild-type strain. All of them mapped to a single locus pstABCDEF operon, suggesting a simple means to limit acetoin toxicity in dairy bacteria and to improve its production.

Keywords: Lactococcus lactis; acetoin; fatty acids; pst operon.

FIG 1

Acetoin biosynthesis pathway in L. lactis MG1363. Als, acetolactate synthase; Ald, acetolactate decarboxylase. Acetoin contains a hydroxyl (R-OH) and a keto (R₁[C═O]R₂) group.

FIG 2

Deletion of recA leads to acetoin sensitivity. Growth curves of (A) the WT strain and (B) a ΔrecA mutant without (circles) and in the presence of (squares) 0.2 M acetoin, under static fermentation in M17Glu1%. Curves are presented in logarithmic scale and representative of three independent experiments.

FIG 3

Acetoin is mutagenic and inhibits DNA polymerase. (A) Cells were grown in M17Glu1% under static growth conditions in the presence of 0.2 M acetoin or not. After overnight growth, cultures were diluted and each suspension was spread on solid medium supplemented with rifampin or not. Data, presented in logarithmic scale, are means with standard deviations from three independent experiments. Statistical significance was determined by unpaired, nonparametric Mann-Whitney tests, as recommended for small sample sizes. *, P  ≤ 0.05. (B) In a PCR mixture acetoin is added at 11, 22, 33 mM (left to right). C, PCR control with no acetoin; M, 1 log DNA ladder. Photo is representative of two independent experiments.

FIG 4

Toxicity of acetoin depends on its keto group. (A) Structure of acetoin and derivatives: diacetyl is produced from spontaneous oxidation of acetolactate or acetoin, whereas 2,3-butanediol by reduction. Structure of diacetyl is similar to methylglyoxal. (B) Growth of a ΔrecA mutant without and with different compounds: no compound, black circle; acetoin, black square; 2,3-butanediol, diamond; diacetyl, triangle. Cells were cultured in M17Glu1%. Each compound was tested at 0.2 M. Curves are presented in logarithmic scale and representative of three independent experiments. Methylglyoxal gave similar results to that of diacetyl ones. (C) PCR assays. Acet, acetoin; Diac, diacetyl, 2,3-But, 2,3-butanediol; MG, methylglyoxal. Experiments are performed as described in Fig. 3.

FIG 5

Fatty acid C18:1 cyclopropanation to cycC19:0 participates in acetoin resistance. (A) The WT strain was cultured in M17Glu0.5% at 30°C and collected at OD₆₀₀ = 0.5 for membrane fatty acid extraction and analysis. 0.2 M acetoin was added at OD₆₀₀ = 0.1. Black bars, no acetoin; gray bars, with acetoin. Results are means with standard deviations from three independent experiments. (B) Overnight cultures of the WT strain and a Δcfa mutant were diluted in M17 broth, and 5 μl of each dilution was loaded on agar M17Glu supplemented with 0.35 M acetoin. After 48 h, bacterial counts were determined. Data are means with standard deviations from three independent experiments. Black bars, no acetoin; gray bars, with acetoin. Statistical significance was determined by unpaired, nonparametric Mann-Whitney tests, as recommended for small sample sizes. *, P ≤ 0.05.

FIG 6

Inactivation of the pst operon enhances acetoin resistance. (A) Locus of pstABCDEF operon. Arrows indicate insertion site of transposon; red, clones isolated under respiration condition; black, clones isolated under aerobic fermentation condition (B) Schema of phosphate transporter in membrane: PstE (also named PstS) and F, lipoproteins; PstC and D, permease; PstA and B, ATPases; Pi, inorganic phosphate. (C) Resistance of a pstA mutant and the WT strain to acetoin. After overnight growth, cells were diluted and 5 μl of each suspension was loaded on M17Glu1% agar plate supplemented with 0.4 M acetoin. Bacterial counts were determined after 48 h of incubation. Data are means with standard deviations from three independent experiments. Black bars, no acetoin; gray bars, with acetoin; white bar, with acetoin and 20 mM phosphate (Pi). Statistical significance was determined by unpaired, nonparametric Mann-Whitney tests, as recommended for small sample sizes *, P ≤ 0.05. ns, not significant.

FIG 7

The pst operon deletion partially rescues ΔclpP and Δcfa mutants against acetoin toxicity but not a ΔrecA mutant. Overnight cultures were tested on M17Glu agar plates as described in Fig. 6. Data are means with standard deviations from three independent experiments. Acetoin was added at 0.35 M in plates. Black bars, no acetoin; gray bars, with acetoin. Statistical significance was determined by unpaired, nonparametric Mann-Whitney tests, as recommended for small sample sizes, *, P ≤ 0.05; ns, not significant.

FIG 8

Acetoin affects the proteome of L. lactis strain MG1363. The WT strain was treated or not to 0.2 M acetoin. Cytosolic proteins were separated according to their isoelectric point (first dimension) then according to their molecular weight (second dimension). Proteins were stained by InstantBlue dye. Identification of proteins was determined from mass spectrometry analysis. Figure is representative of two independent experiments (see Materials and Methods). Phosphoproteome analysis is described in the supplemental material.