Citrate Metabolism by Enterococcus faecalis FAIR-E 229

Abstract

Citrate metabolism by Enterococcus faecalis FAIR-E 229 was studied in various growth media containing citrate either in the presence of glucose or lactose or as the sole carbon source. In skim milk (130 mM lactose, 8 mM citrate), cometabolism of citrate and lactose was observed from the first stages of the growth phase. Lactose was stoichiometrically converted into lactate, while citrate was converted into acetate, formate, and ethanol. When de Man-Rogosa-Sharpe (MRS) broth containing lactose (28 mM) instead of glucose was used, E. faecalis FAIR-E 229 catabolized only the carbohydrate. Lactate was the major end product, and small amounts of ethanol were also detected. Increasing concentrations of citrate (10, 40, 70, and 100 mM) added to MRS broth enhanced both the maximum growth rate of E. faecalis FAIR-E 229 and glucose catabolism, although citrate itself was not catabolized. Glucose was converted stoichiometrically into lactate, while small amounts of ethanol were produced as well. Finally, when increasing initial concentrations of citrate (10, 40, 70, and 100 mM) were used as the sole carbon sources in MRS broth without glucose, the main end products were acetate and formate. Small amounts of lactate, ethanol, and acetoin were also detected. This work strongly supports the suggestion that enterococcal strains have the metabolic potential to metabolize citrate and therefore to actively contribute to the flavor development of fermented dairy products.

The industrial importance of lactic acid bacteria is mainly based on the ability of these organisms to rapidly ferment carbohydrates and to convert them into lactic acid and, to a lesser degree, into other flavor compounds. Lactic acid provides protection against spoilage by nonacidophilic organisms. On the other hand, many lactic acid bacteria are also able to ferment a number of noncarbohydrates, including citrate. Citrate metabolism plays an important role in many food fermentations involving lactic acid bacteria, since it occurs in many natural substrates, such as milk, vegetables, and fruits (21). The behavior of lactic acid bacteria may differ from one species to another, and not all lactic acid bacteria are able to metabolize citrate (24).

The ability to metabolize citrate is invariably linked to endogenous plasmids that contain the gene encoding the transporter which is responsible for citrate uptake from the medium (2). Since citrate is a highly oxidized substrate, no reducing equivalents, such as NADH, are produced during its degradation, which results in the formation of metabolic end products other than lactic acid. Some of these end products, such as diacetyl, acetaldehyde, and acetoin, have very distinct aroma properties and significantly influence the quality of fermented foods (20). For instance, diacetyl determines the aromatic properties of fresh cheese, fermented milk, cream, and butter (12) but is considered the most important off-flavor compound in the brewing process and in the wine industry (22). The breakdown of citrate also results in the production of carbon dioxide, which can contribute to the texture of some fermented dairy products (25).

Strains of Lactococcus lactis and Leuconostoc sp. have been extensively studied with respect to citrate metabolism and production of aroma compounds (2, 7, 13, 21, 25, 33). In contrast, only limited data concerning citrate metabolism by Enterococcus strains are available (9, 16, 37).

It should be noted that significant numbers of enterococci are present in many dairy products, especially those originating from the Mediterranean area. In many cheeses, such as Comté, Cebreiro, Mozzarella, Kefalotyri, Serra, Manchego, Feta, and Teleme, enterococci comprise a major part of the fresh cheese microflora, and in some cases they are the predominant microorganisms in the fully ripened product (3, 4, 8, 27, 28, 30, 36). It has been concluded in many reports that enterococci may have an important role in cheese production, contributing to the ripening and quality of the mature products (5, 30, 34, 35). Moreover, some researchers have suggested that enterococci may play a role in the development of the aroma and flavor of many cheeses, probably due to citrate catabolism and lipolysis (4, 10, 18, 34).

The aim of the present study was to examine citrate metabolism by the Enterococcus faecalis FAIR-E 229 strain, isolated from Cheddar cheese, in various growth media containing citrate either in the presence of glucose or lactose or as the sole energy source.

MATERIALS AND METHODS

Strain.

The microorganism used throughout this study was E. faecalis FAIR-E 229 (Dairy Products Research Centre, Teagasc, Moorepark, Fermoy, Ireland), which was isolated from Cheddar cheese. Detailed strain information is available in the Catalogue of Enterococci of the FAIR-E Collection; this collection is maintained at the BCCM/LMG Bacteria Collection, Laboratory of Microbiology, University of Ghent, Ghent, Belgium (38). The strain was stored at −80°C in de Man-Rogosa-Sharpe (MRS) broth (Oxoid) containing 25% (vol/vol) glycerol (Sigma). Before experimental use the strain was propagated twice in the appropriate medium (see below) at 37°C for 24 h.

Growth conditions.

Ten media were used for growth of E. faecalis FAIR-E 229. First, the strain was grown in skim milk (10%, wt/vol) supplemented with 0.3% (wt/vol) yeast extract (Oxoid) (medium M₁) and in MRS broth containing 28 mM lactose (Merck) instead of glucose (medium M₂). In a second step, MRS broth preparations without acetate but with different concentrations of citrate (10, 40, 70, and 100 mM; Merck) were used as the growth media (media M₃, M₄, M₅, and M₆, respectively). Finally, growth was examined in MRS broth which lacked both glucose and acetate but contained different concentrations of citrate (10, 40, 70, and 100 mM) (media M₇, M₈, M₉, and M₁₀, respectively). The pH values of the media other than the skim milk medium were adjusted to approximately 6.2 prior to sterilization. Growth was carried out microaerophilically at 37°C for 48 h at an uncontrolled pH. Experiments were performed in duplicate and repeated if the experimental variation exceeded 5%.

Growth of E. faecalis FAIR-E 229 was assessed by measuring the optical density at 600 nm (OD₆₀₀) (synthetic media) or as described by Kanasaki et al. (23) (skim milk medium) and by measuring the pH (632 pH meter; Metrohm Herisau, Herisau, Switzerland). The maximum specific growth rate (μ_max) was determined by linear regression (as indicated by the correlation coefficient [r²]) from plots of ln optical density/optical density at zero time versus time.

Analysis of metabolites.

Samples were taken over a period of 48 h. The glucose, lactose, citrate, lactate, acetate, and formate contents of culture supernatants were determined by high-performance liquid chromatography analysis (Varian Associates Inc., Palo Alto, Calif.). First, cells were removed by centrifugation (5,000 × g, 15 min, 4°C; Heraeus Sepatech Biofuge 22R). A 20-μl sample of the culture supernatant was injected into an Aminex HPX-87H column (300 by 7.8 mm; Bio-Rad, Hercules, Calif.) connected to a refractive index detector (model LC 1240; GBC Scientific Equipment Pty. Ltd., Dandenong, Victoria, Australia). Elution was performed at 35°C (or at 60°C when both citrate and glucose were present in the medium) with 5 mM H₂SO₄ at a flow rate of 0.5 ml/min. Data were collected and analyzed by using a 746 data module (Waters Corporation, Milford, Mass.).

Ethanol and acetoin were isolated, and the ethanol and acetoin contents were determined with a dynamic headspace analyzer (HS-40; Perkin-Elmer, Ueberlingen, Germany) coupled to a QP 5050 gas chromatography-mass spectrometry system (Shimadzu Scientific Instruments, Inc., Columbia, Md.). Five-milliliter portions of culture supernatant (see above) were transferred into 20-ml vials, and the vials were sealed. The samples were incubated at 75°C for 15 min, purged, and pressurized with 35 ml of ultrapure helium gas per min. The isolated volatile compounds were driven through the transfer line (thermostat temperature, 90°C) and injected into an HP INNOWax capillary column (60 m by 0.25 mm) that was coated with cross-linked polyethylene glycol (film thickness, 0.25 μm) and connected without splitting to the ion source of a QP 5050 quadrupole mass spectrometer (interface line temperature, 250°C) operating in the scan mode with a mass range of m/z 40 to 300 at a rate of 1 scan/s. The carrier gas was helium (flow rate, 0.6 ml/min), and the injector temperature was set at 200°C. The temperature program was as follows: 35°C for 3 min, increase to 80°C at a rate of 5°C/min; 80°C for 3 min; and then increase to 200°C at a rate of 8°C/min. Compounds were identified by computer matching of mass spectral data with data in the Shimadzu NIST62 mass spectral database and by comparing the retention times and mass spectra to those of standard compounds (Sigma). Quantification was performed by integrating the peak areas of total ion chromatograms with the Shimadzu Class 500 software and using the appropriate standard curves.

The concentrations of both water-soluble and volatile metabolites were expressed as millimolar concentrations, and stoichiometric calculations were performed for all samples taken during growth.

RESULTS

The results are summarized in Tables 1 and 2, in which exact concentration data are given for the 48-h samples. The stoichiometry determined for the initial and final products was valid for growth in all of the media tested.

TABLE 1.

Growth characteristics and metabolic profiles of E. faecalis FAIR-E 229 in media containing both citrate and glucose (or lactose) as carbon sources after 48 h of incubation at 37°C

Mediuma	Final pH	ΔpH	OD600	μmax (h−1)b	Glucose or lactose consumption (mM)	Citrate consumption (mM)	Lactate production (mM)	Ethanol production (mM)	Acetate production (mM)	Formate production (mM)
M1	4.80	1.56	1.92	0.53	12.8	3.1	48.7	2.7	3.5	3.1
M2	4.74	1.46	1.33	0.66	12.9	0.0	50.3	1.5	0.0	0.0
M3	3.98	1.97	1.29	0.56	31.5	0.0	61.7	1.0	0.0	0.0
M4	4.12	1.80	1.66	0.80	48.2	0.0	95.9	1.5	0.0	0.0
M5	4.25	1.61	1.84	0.88	60.5	0.0	119.6	1.8	0.0	0.0
M6	4.36	1.48	1.85	0.76	71.6	0.0	140.0	3.0	0.0	0.0

^aMedium M₁, 10% skim milk supplemented with 0.3% yeast extract; medium M₂, MRS broth containing lactose (28 mM) instead of glucose; medium M₃, MRS broth lacking acetate but containing 10 mM citrate; medium M₄, MRS broth lacking acetate but containing 40 mM citrate; medium M₅, MRS broth lacking acetate but containing 70 mM citrate; medium M₆, MRS broth lacking acetate but containing 100 mM citrate. 

^bμ_max, maximum specific growth rate. 

TABLE 2.

Growth characteristics and metabolic profiles of E. faecalis FAIR-E 229 in media containing citrate as the sole carbon source after 48 h of incubation at 37°C

Mediuma	Final pH	ΔpH	OD600	μmax (h−1)b	Citrate consumption (mM)	Acetate production (mM)	Formate production (mM)	Lactate production (mM)	Ethanol production (mM)	Acetoin production (mM)
M7	6.30	0.00	0.33	0.83	12.6	18.5	6.2	5.6	1.8	0.0
M8	6.48	0.24	0.59	0.71	41.1	72.8	28.0	4.5	2.3	0.4
M9	6.54	0.35	0.64	0.55	71.4	128.3	45.3	7.4	4.5	0.9
M10	6.79	0.58	1.00	0.49	94.1	168.0	52.5	12.5	5.5	1.4

^aMedium M₇, MRS broth lacking glucose and acetate but containing 10 mM citrate; medium M₈, MRS broth lacking glucose and acetate but containing 40 mM citrate; medium M₉, MRS broth lacking glucose and acetate but containing 70 mM citrate; medium M₁₀, MRS broth lacking glucose and acetate but containing 100 mM citrate. 

^bμ_max, maximum specific growth rate. 

When skim milk supplemented with yeast extract (medium M₁; 130 mM lactose, 8 mM citrate) was used as the substrate, cometabolism of lactose and citrate took place. Lactose and citrate were catabolized simultaneously from the beginning of growth. The major end product detected was lactate, which was derived exclusively from degradation of lactose. Small amounts of acetate, formate, and ethanol were produced as well, which were stoichiometrically justified by the catabolism of citrate (Table 1 and Fig. 1). In this natural medium, E. faecalis FAIR-E 229 grew well and had a maximum growth rate of 0.53 h⁻¹. After 48 h of growth, the change in pH (ΔpH) was approximately 1.56.

FIG. 1.

Schematic pathway showing the metabolic relationships between citrate and glucose. 1, citrate lyase; 2, oxaloacetate decarboxylase; 3, lactate dehydrogenase; 4, acetolactate synthase; 5, acetolactate decarboxylase; 6, diacetyl/acetoin reductase; 7, pyruvate dehydrogenase complex; 8, pyruvate formate lyase; 9, acetate kinase; 10, alcohol dehydrogenase.

In contrast to the results obtained with skim milk, in MRS broth containing 28 mM lactose instead of glucose (medium M₂), E. faecalis FAIR-E 229 catabolized only the lactose, and lactate was the main end product (97.2%) (Table 1). Small amounts of ethanol, also derived from lactose catabolism, were produced, while citrate was not catabolized. The maximum growth rate was 0.66 h⁻¹, and the ΔpH after 48 h of growth was approximately 1.46.

Likewise, when both glucose and citrate were used as carbon sources (media M₃, M₄, M₅, and M₆), only glucose was catabolized, and citrate was not catabolized. Glucose was stoichiometrically converted into lactate, which was the major end product (98.3%). Small quantities of ethanol, derived from glucose catabolism, were also produced (Table 1 and Fig. 1). Figure 2 shows the growth of E. faecalis FAIR-E 229 and the kinetics of glucose catabolism in medium M₆ (in the presence of 100 mM citrate). It is interesting that in the presence of different initial concentrations of citrate (10, 40, 70, and 100 mM; media M₃, M₄, M₅, and M₆), glucose consumption and thus lactate production were enhanced in a linear way (Fig. 3A). In all media, the pH declined during growth, but the final pH values increased as the initial citrate concentration increased (Fig. 3A). Finally, it was observed that as the citrate concentration was increased up to 70 mM, the maximum growth rate and the final OD₆₀₀ and ethanol production values increased. In the presence of 100 mM citrate, the OD₆₀₀ remained unchanged, the maximum growth rate decreased slightly, and ethanol production increased (Table 1 and Fig. 3A).

FIG. 2.

Growth and metabolite kinetics of E. faecalis FAIR-E 229 in medium M₆ (MRS broth without acetate supplemented with 100 mM citrate) at 37°C. Symbols: ●, OD₆₀₀; ■, pH; ▴, glucose concentration; ▾, lactate concentration; ⧫, ethanol concentration.

FIG. 3.

(A) Lactate production, maximum specific growth rate (μ_max), and pH as a function of the initial citrate concentration in media M₃, M₄, M₅, and M₆ after growth of E. faecalis FAIR-E 229 at 37°C for 48 h. Symbols: ●, lactate concentration; ■, maximum specific growth rate; ▴, pH. (B) Acetate production, maximum specific growth rate, and pH as a function of the initial citrate concentration in media M₇, M₈, M₉, and M₁₀ after growth of E. faecalis FAIR-E 229 at 37°C for 48 h. Symbols: ●, acetate concentration; ■, maximum specific growth rate; ▴, pH.

When citrate was used as the sole carbon source at different initial concentrations (10, 40, 70 and 100 mM; media M₇, M₈, M₉, and M₁₀), citrate was stoichiometrically converted to acetate and formate and, to a lesser degree, to lactate, ethanol, and acetoin (Table 2 and Fig. 1). Figure 4 shows the growth of E. faecalis FAIR-E 229 and the kinetics of citrate catabolism in medium M₁₀ (100 mM citrate). It is obvious that growth and metabolic activity ceased when the citrate was exhausted. It is interesting that as the initial concentration of citrate increased (10, 40, 70, and 100 mM; media M₇, M₈, M₉, and M₁₀), acetate production increased and the maximum growth rate declined, both in a linear way (Fig. 3B). Furthermore, in all media the pH decreased during the early exponential phase and then steadily increased until the end of growth, and the final pH values increased as the initial citrate concentration increased (Fig. 3B).

FIG. 4.

Growth and metabolite kinetics of E. faecalis FAIR-E 229 in medium M₁₀ (MRS broth without glucose and acetate supplemented with 100 mM citrate) at 37°C. Symbols: ●, OD₆₀₀; ■, pH; ▴, citrate concentration; ▾, acetate concentration; ⧫, formate concentration.

DISCUSSION

Citrate metabolism by lactic acid bacteria is essential in many fermented foods and beverages, since many compounds with special flavor properties are produced. However, numerous authors have reported that citrate metabolism does not support growth in some lactic acid bacteria. This conclusion was based on the inability of the lactic acid bacteria to grow in batch cultures to which excess citrate was added as the sole energy source (6, 7). Sometimes, lactic acid bacteria need sugars, such as glucose or lactose, as cosubstrates in order to consume citrate (7, 17, 19).

When E. faecalis FAIR-E 229 was grown in skim milk (medium M₁), lactose and citrate were cocatabolized. Lactose was converted to lactate, and citrate was converted to acetate, formate, and ethanol. Cometabolism of lactose and citrate in milk has been reported previously (14, 16, 32). Raffe (32) suggested that E. faecalis strains produced lactate from lactose and acetate from citrate when they were grown in skim milk. Furthermore, El Attar et al. (14) reported that L. lactis strains grown in skim milk exhausted almost all of the citrate available and 25% of the lactose, producing mainly lactate along with formate, ethanol, acetate, CO₂, and acetoin. Finally, Freitas et al. (16) concluded that E. faecalis and Enterococcus faecium strains cultured in ovine and caprine milk produced large amounts of lactate, followed by formate and acetate, suggesting indirectly that lactose and citrate cometabolism occurs.

On the other hand, in MRS broth containing 28 mM lactose instead of glucose (medium M₂), E. faecalis FAIR-E 229 catabolized only the lactose, producing lactate as the main end product and small amounts of ethanol. It is noteworthy that although in medium M₂ the lactose concentration was considerably lower than the lactose concentration in skim milk (medium M₁), the same amount of lactose was consumed in both media, whereas citrate was not catabolized at all in medium M₂. Whether lactose concentration has an effect on citrate catabolism cannot be determined at this time; further experiments are needed to answer this question.

Reports on the energetics of citrate metabolism are contradictory. Some authors claimed that in carbohydrate-containing media citrate was cometabolized and at the same time stimulated growth of Lactobacillus plantarum (24), L. lactis subsp. lactis biovar diacetylactis (25), and Lactobacillus amylovorus (39). In these cases, the main end products observed were lactate, acetate, formate, and ethanol. On the other hand, it has been shown that growth of L. lactis subsp. lactis biovar diacetylactis strains was not enhanced by citrate (15). Likewise, Palles et al. (31) suggested that citrate did not affect the growth rate of Lactobacillus casei or Lactobacillus plantarum when it was cometabolized with glucose or galactose.

In the present study, increasing the citrate concentration enhanced the growth rate of E. faecalis FAIR-E 229 in media containing both glucose and citrate. Glucose consumption and lactate production were also stimulated. Glucose was stoichiometrically converted mainly to lactate and small amounts of ethanol. Kimoto et al. (25) reported similar stimulation of both growth and glucose consumption in L. lactis subsp. lactis biovar diacetylactis strains in the presence of increasing citrate concentrations, but at the same time there was cometabolism of glucose and citrate. However, in our study citrate was not catabolized at all in the presence of glucose. Likewise, cometabolism was not observed in Lactobacillus rhamnosus (11, 12).

In all media containing both glucose and citrate, the pH declined during growth, but the final pH increased as the initial citrate concentration increased. This may be attributed mainly to the buffering capacity of the noncatabolized citrate. Similar results for the effect of citrate on pH have been reported for citrate metabolism by L. rhamnosus (11). Additionally, the production of ethanol indicates that there is decarboxylation of pyruvate, derived from glucose, to acetyl coenzyme A (acetyl-CoA) via the pyruvate decarboxylase complex (Fig. 1). The carbon dioxide formed may also contribute, to a lesser degree than citrate, to the higher pH values, as previously suggested (13).

As previously reported for other lactic acid bacteria (11, 12, 21, 24, 33, 37), E. faecalis FAIR-E 229 was able to grow well in media containing citrate as the sole carbon source and produced acetate and formate as the main end products. This suggests that citrate acted as an energy source. The energy was generated mostly from the conversion of acetyl-CoA to acetate, meaning that citrate acted as an electron acceptor, which resulted in greater production of acetate and ATP, probably via the acetate kinase pathway (Fig. 1). However, there was not enough ATP to account for the observed increase in biomass as a result of the higher citrate concentrations. Apparently, additional energy is produced during the initial conversion of citrate into pyruvate (20). Furthermore, recent studies performed with L. lactis subsp. lactis biovar diacetylactis (2, 21) and Leuconostoc oenos (29) indicated that uptake of citrate was coupled to generation of a proton motive force, which was shown to be strong enough to drive the additional ATP synthesis. This is in accordance with our results, since the major end product in media M₇, M₈, M₉, and M₁₀ was acetate. Conversion of citrate, used as the sole carbon source, mainly to acetate has been also reported for L. plantarum (24), L. lactis subsp. lactis biovar diacetylactis (21), and L. rhamnosus (12). Furthermore, similar results have been reported for E. faecalis (37) and E. faecium (9).

Besides acetate, large quantities of formate were also detected in the present study. The microaerophilic conditions and the pH values higher than 6.0 that prevailed during growth of E. faecalis FAIR-E 229 favored regulation of the pyruvate formate lyase (PFL), which leads to formation of formate and acetyl-CoA. The PFL system is sensitive to oxygen and active at medium pH values higher than 6.0, as reported for E. faecalis and Streptococcus mutans (1, 26).

The large amounts of acetate and formate produced indicated that most probably the PFL system, the pyruvate dehydrogenase complex, and acetate kinase were the predominant enzyme systems expressed during growth of E. faecalis FAIR-E 229 in media containing citrate as the sole energy source. On the other hand, the small quantities of lactate, acetoin, and ethanol detected suggested that lactate dehydrogenase, acetolactate synthase, and alcohol dehydrogenase exhibited relatively low activities, since no energy in the form of ATP was produced via these pathways (Fig. 1).

In media M₇, M₈, M₉, and M₁₀, the pH decreased during the early exponential phase and then steadily increased until the end of growth, and the final pH values increased as the initial citrate concentration increased. This phenomenon can be explained by the formation of carbon dioxide. According to the pathways presented in Fig. 1, the theoretical ratio of the acids produced (acetate, formate, and lactate) to carbon dioxide is approximately 2.5:1. Carbon dioxide in the form of the weak acid carbonate (pK_a 6.1) buffers the acidity of acetate (pK_a 4.76), formate (pK_a 3.75), and lactate (pK_a 3.86).

Citrate present in milk at concentrations of 8 to 9 mM may be metabolized or cometabolized during cheese manufacture by several strains of lactic acid bacteria, including enterococci. This metabolic process leads to the formation of a number of minor products important for the organoleptic properties of cheese (19). The present report clearly shows that although E. faecalis FAIR-E 229 cannot catabolize citrate in the presence of glucose, it has the ability to metabolize citrate in milk in the presence of lactose. Moreover, this strain is able to use citrate as a sole carbon source for growth and energy production. Citrate metabolism by enterococcal strains is a significant finding, since this group comprises a major part of the microflora in several types of cheese and may thus contribute to the distinct flavor properties of the cheeses.