Effect of Different NADH Oxidase Levels on Glucose Metabolism by Lactococcus lactis: Kinetics of Intracellular Metabolite Pools Determined by In Vivo Nuclear Magnetic Resonance

Organisms used and plasmid construction.

L. lactis MG1363 and two derivative strains were used throughout this study. The NOX⁺ strain was obtained by transformation of L. lactis NZ3900 with plasmid pNZ2600, a derivative of pNZ8020 containing the water-forming NADH oxidase encoded by the nox-2 gene of S. mutans under control of the lactococcal nisA promoter (27). Strain NZ3900 is an MG1363 derivative which contains the nisRK genes integrated in the chromosome, which allows nisin-controlled expression of the nox-2 gene (12). The NOX⁻ strain is a derivative of MG1363 carrying a deletion of nox-2 (or the noxE gene of L. lactis). Cloning of the nox-2 gene of L. lactis MG1363, which encodes the major water-forming NADH oxidase, was recently described (20) (GenBank accession no. AY046926). The chromosomal nox-2 gene was knocked out in L. lactis MG1363 by using a double-crossover gene disruption strategy. The 3′ and 5′ flanking regions required for homologous recombination into the lactococcal chromosome were obtained by digestion of pUC-5′NOX (20) with EcoRI and NspV, which generated a 1.1-kb fragment containing the upstream and 5′ coding region of nox-2, and by digestion of pUC-3′NOX (20) with EcoRI and SspI, which generated a 1.0-kb fragment containing the downstream and 3′ coding region of nox-2. These two fragments were cloned into HincII-NarI-digested pUC19 by using three-point ligation. The resulting plasmid contained the entire nox-2 gene (the 5′ and 3′ ends were fused at the EcoRI site) and was designated pUC-NOX. The erythromycin resistance gene was isolated as a PstI-AccI fragment from pUC19Ery (43) and was blunted by using T4 polymerase. This erythromycin resistance cassette was cloned into the EcoRI site of pUC-NOX after the EcoRI sticky ends were blunted by using Klenow polymerase. The resulting plasmid, containing the erythromycin resistance gene in the same orientation as the nox-2 gene, was transformed into competent L. lactis MG1363 cells, and erythromycin-resistant colonies were selected. In 15 of these colonies the organization of the chromosomal nox-2 locus was analyzed by Southern blotting, and two colonies that contained the desired double crossover nox-2::ery gene disruption were selected. One of these mutant strains (designated NZ9020) was selected and used in this study.

Growth conditions.

Strains were grown in a 2-liter fermentor (B. Braun Biostat MD) in chemically defined medium (34) containing glucose (10 g · liter⁻¹) at 30°C and pH 6.5 or 5.5. For growth of the NOX⁺ and NOX⁻ strains, the medium was supplemented with chloramphenicol (10 mg · liter⁻¹) and erythromycin (5 mg · liter⁻¹), respectively. For overproduction of NADH oxidase by the NOX⁺ strain, nisin (5 μg · liter⁻¹) was added to the medium when an optical density at 600 nm (OD₆₀₀) of 0.5 was reached, and cells were harvested in the mid-log growth phase (OD₆₀₀, 2.2). Anaerobiosis was attained by flushing sterile argon through the medium in the fermentor for 1 h before inoculation. Under aerobic conditions, dissolved oxygen was monitored with a polarographic oxygen electrode (Ingold). The electrode was calibrated to zero and 100% by bubbling sterile argon and air, respectively, through the medium. A specific air tension of 90% was maintained by automatic control of the airflow.

In vivo NMR experiments.

Cells were harvested in the mid-log growth phase (OD₆₀₀, 2.2), centrifuged, washed twice, and suspended in 50 mM potassium phosphate (KP_i) buffer (pH 6.5, 5.5, or 4.5) to a protein concentration of approximately 13 mg · ml⁻¹ (equivalent to 22 mg [dry weight] · ml⁻¹), and the suspension was placed in a bioreactor (50 ml). Anaerobic conditions were established as described previously (32), but the thickness of the tubing was increased to 0.8 mm. For experiments under an O₂ atmosphere, a micro pO₂ electrode (Lazar Research Laboratories, Inc.) was included in the experimental setup described previously (32). To ensure an adequate level of oxygenation, an air lift system (37) with pure oxygen was used inside the NMR tube, and in addition, oxygen was continuously bubbled through the cell suspension in the bioreactor; the rate of circulation of the cell suspension between the fermentor and the NMR tube was 36 ml · min⁻¹.

In vivo determination of NAD⁺ and NADH by ¹³C-NMR was carried out with cell suspensions grown in chemically defined medium without aspartate and asparagine and supplemented with nicotinic acid (5 mg · liter⁻¹) specifically labeled on carbon 5 as described by Neves et al. (31). The NOX⁺ strain was unable to grow in this medium.

After acquisition of an initial spectrum, [1-¹³C]glucose or [6-¹³C]glucose (20 mM) was supplied, and the time courses for glucose consumption, product formation, and buildup and decline of intracellular metabolite pools were monitored. The use of [6-¹³C]glucose allowed monitoring of the resonances of intracellular glucose 6-phosphate (G6P) that otherwise were obscured by the intense resonances due to [1-¹³C]glucose. Identical intracellular pools were observed for all metabolites other than G6P regardless of the position of the label in glucose. After glucose exhaustion and when no changes in the resonances due to end products and intracellular metabolites were observed, an aliquot of the cell suspension was passed through a French press; the resulting cell extract was incubated at 80°C for 10 min in a stoppered tube and cooled on ice, and cell debris and denatured macromolecules were removed by centrifugation. The supernatant, designated the NMR sample extract, was used for quantification of end products and minor metabolites.

Although the results of individual experiments are shown below, NMR experiments were performed at least twice for each type of experimental conditions and were highly reproducible. ¹³C spectra were acquired with a Bruker DRX500 spectrometer at 30°C by using a quadruple-nucleus probe head (32). Carbon chemical shifts were referenced to the resonances of external methanol designated at 49.3 ppm.

Quantification of end products.

Lactate, acetoin, acetate, 2,3-butanediol, and formate were quantified in NMR sample extracts by ¹H-NMR (32). The concentrations of minor products (e.g., pyruvate, ethanol, and diacetyl) and the concentrations of metabolites present at the end of the in vivo NMR experiments (phosphoenolpyruvate [PEP] and 3-phosphoglycerate [3-PGA]) in NMR sample extracts were determined by ¹³C analysis as previously described (33).

Quantification of intracellular metabolites in living cells by ¹³C-NMR.

Due to the fast pulsing conditions used to acquire in vivo ¹³C spectra, a direct correlation between concentrations and peak intensities could be established only by using correction factors. The correction factors for C-1 and C-6 of fructose 1,6-bisphosphate [Fru(1, 6)P₂] (0.73 ± 0.03) and for C-3 of 3-PGA and PEP (0.71 ± 0.01) were determined as described by Neves et al. (33). For C-6 of G6P, the correction factor (0.63 ± 0.04) was determined from spectra of an NMR sample extract to which G6P (30 mM) was added; the extract was circulated through the NMR tube at a rate similar to that used for suspensions of living cells. The quantitative kinetic data for intracellular metabolites were calculated from the areas of the corresponding resonances by using the correction factors and comparing the intensities with the intensity of the lactate resonance in the last spectrum of the sequence.

Metabolite concentrations were calculated by using a value of 2.9 μl · mg of protein⁻¹ for the intracellular volume (35). The concentration limit for detection of intracellular metabolites under the conditions used to acquire spectra of living cells (acquisition time, 30 s) was approximately 4 mM.

Enzyme activity measurements.

All enzymes were assayed at 30°C by using freshly prepared cell extracts. The extracts were obtained by disruption of cell suspensions (prepared as described above for NMR experiments; OD₆₀₀, 2.2) in a French press (three passages, 120 MPa), followed by centrifugation at 30,000 × g. The protein concentration was determined by the method of Bradford (5). l-Lactate dehydrogenase (LDH) (EC 1.1.1.27) and pyruvate kinase (PK) (EC 2.7.1.40) activities were measured as described previously (15). Glyceraldehyde 3-phosphate dehydrogenase (Gra-3-P dehydrogenase) (EC 1.2.1.12) was assayed as described by Even et al. (13). 6-Phosphofructokinase (PFK) (EC 2.7.1.11) activity was measured by the method of Fordyce et al. (14).

To measure enolase activity, a modification of the assay described by Schäfer and Schönheit (38) was used; the 1-ml reaction mixtures contained 100 mM Tris-HCl buffer (pH 7.2), 5 mM MgCl₂, 5 mM ADP, 0.3 mM NADH, 10 U of PK (rabbit muscle), and 10 U of LDH (pig heart muscle), and 2-PGA (3 mM) was used to initiate the reaction.

NADH peroxidase activity was measured under anaerobic conditions by using a modification of the assay described by Zitzelsberger et al. (44). The 1-ml assay mixture contained 50 mM Tris-HCl (pH 7.8), 0.3 mM NADH, and 10 mM H₂O₂. SOD (EC 1.15.1.1) activity was measured by using the protocol of Beauchamp and Fridovich (2), in which a xanthine-xanthine oxidase system is used to generate O₂⁻ and nitroblue tetrazolium is used as an indicator. One unit of SOD activity was defined as the amount of SOD that resulted in 50% inhibition of the reduction of nitroblue tetrazolium. Total NADH oxidase activity (sum of H₂O-forming NADH oxidase, H₂O₂-forming NADH oxidase, and NADH peroxidase activities) was measured as described by Lopez de Felipe et al. (27). H₂O₂-forming NADH-dependent activity was determined by measuring the amount of H₂O₂ formed in 1-ml reaction mixtures containing 50 mM KP_i buffer (pH 7.0), 0.3 mM EDTA, and 0.3 mM NADH. The H₂O₂ formed was measured colorimetrically at 600 nm by the method described by Meiattini (30). The activity of H₂O-forming NADH oxidase was calculated from the difference between the total NADH oxidase activity and the H₂O₂-forming NADH-dependent activity and was corrected for NADH peroxidase activity.

Purification of NADH oxidase.

L. lactis NOX⁺ cells (34 g, wet weight) were harvested in the mid-log growth phase, washed twice with 5 mM KP_i buffer (pH 6.5), and suspended in the same buffer containing 0.5 mM phenylmethylsulfonyl fluoride, 2 μg of antipain per ml, and 2 μg of leupeptin per ml. A crude extract was prepared by passage through a French press and ultracentrifugation (5 h at 150,000 × g). The resulting supernatant was concentrated by ultrafiltration (Diaflo YM3; Amicon Corporation). Purification was performed by the method described by Higuchi et al. (18) by using a first step consisting of precipitation with ammonium sulfate and four subsequent chromatographic steps (DEAE-Sepharose, Resource Q, phenyl-Sepharose HP, and Resource Q). All the purification steps were carried out in the presence of protease inhibitors. The enzyme preparation was determined to be pure by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The NADH oxidase activity assay described above was used to monitor the presence of the enzyme during the purification procedure. The protein concentration was determined from the A₂₇₁ by using the molar extinction coefficient ɛ₂₇₁ = 69.8 mM⁻¹ · cm⁻¹, which was calculated from quantification of amino acid residues in a Pico · Tag amino acid analysis system after acid hydrolysis and derivatization with phenylisothiocyanate.

Determination of the affinity of NADH oxidase for oxygen.

To determine the K_m of NADH oxidase for oxygen, oxymyoglobin was used as an oxygen probe (3). Oxymyoglobin was prepared from horse myoglobin as described by Jünemann et al. (25) and was diluted to a concentration of approximately 14.5 μM in 50 mM KP_i buffer (pH 7.0)-1 mM EDTA saturated with 1% (vol/vol) oxygen in argon. All other solutions used were made anaerobic by sparging with argon. A quartz cuvette was flushed with argon and filled with 2.5 ml of the oxymyoglobin solution, and NADH (1.2 mM) was added. The reaction was started by injecting the purified NADH oxidase (22.7 nM), and visible spectra were acquired sequentially. From the difference in the absorbance values at 580 and 562 nm and the dissociation constant for the reaction of myoglobin with oxygen (K_d = 1.2 μM), the mean concentration of free dissolved oxygen was determined. The rate of oxygen consumption was determined from the time dependency of the total oxygen concentration. Microsoft excel Solver was used to determine the kinetic parameters (K_m and V_max) from the plots of rate of oxygen consumption versus oxygen concentration.

Chemicals.

[1-¹³C]glucose and [6-¹³C]glucose (99% ¹³C enrichment) were obtained from Campro Scientific (Veenendaal, The Netherlands) and ISOTEC (Miamisburg, Ohio), respectively. Horse heart myoglobin was obtained from Sigma Chemical Co. (St. Louis, Mo.). Chromatographic materials were supplied by Pharmacia Fine Chemicals (Uppsala, Sweden). All other chemicals were reagent grade.

Glucose metabolism by the wild-type strain: effect of oxygen.

In vivo NMR was used to study glucose metabolism at pH 6.5 by nongrowing cell suspensions of L. lactis MG1363 under aerobic and anaerobic conditions. Figure 1 shows the kinetics of glucose consumption, end product formation (Fig. 1A and B), and evolution of intracellular metabolite pools (Fig. 1C and D) under aerobic (100% oxygen) and anaerobic conditions. In the absence of oxygen the glucose consumption rate (0.41 ± 0.01 μmol · min⁻¹ · mg of protein⁻¹) was 1.6-fold higher than the glucose consumption rate under aerobic conditions (0.25 ± 0.03 μmol · min⁻¹ · mg of protein⁻¹).

With respect to intracellular metabolites, the G6P pool was maximal within the first few seconds after glucose addition and decreased to undetectable levels parallel with the decrease in the glucose level. Meanwhile, the concentration of Fru(1,6)P₂ reached maxima of 48.8 ± 0.4 mM (anaerobic conditions) and 45.5 ± 0.2 mM (aerobic conditions). The decline in the Fru(1,6)P₂ level was significantly steeper under aerobic conditions, under which it was accompanied by increases in the 3-PGA and PEP concentrations, which reached maximal levels of 33 ± 3 and 13 ± 2 mM, respectively. Under anaerobic conditions, the concentrations of these compounds were lower, 8 ± 2 and 3 ± 1 mM, respectively. The low PEP concentration (close to the in vivo detection limit) hampered reliable quantification in vivo; therefore, the concentration of PEP was determined by using NMR sample extracts.

Under aerobic conditions, lactate (final concentration, 31.4 ± 1.1 mM) was the major product, but about 16% of the carbon flux was channeled towards the production of acetate (6.7 ± 0.7 mM). Under anaerobic conditions, glucose was almost completely metabolized to lactate (37.4 ± 0.3 mM), and acetate (0.5 ± 0.1 mM) accounted only for 1.3% of the carbon from glucose. Other minor products that were detected were ethanol and formate under anaerobic conditions, acetoin under aerobic conditions, and 2,3-butanediol under both conditions.

To rationalize the metabolic data, relevant enzymatic activities were measured in cell extracts of cultures grown under aerobic or anaerobic conditions (Table 1). The activities of the glycolytic enzymes encoded by the las operon, PFK, PK, and LDH, as well as the activity of enolase, decreased significantly in cells grown under aerobic conditions, and the change in LDH activity was especially noticeable; in contrast, the activity of Gra-3-P dehydrogenase increased slightly. The possibility of a direct effect of oxygen on the enzyme activities could be eliminated since the same results were obtained regardless of whether aerobic or anaerobic conditions were used for the assay (data not shown). NADH peroxidase activity, H₂O₂-forming activity dependent on NADH, and H₂O-forming NADH oxidase activity were induced 2.5-, 2- and 3.8-fold, respectively, in cells grown under aerobic conditions, whereas SOD activity was not affected.

Characterization of glucose metabolism in a strain overproducing NADH oxidase.

The total NADH oxidase activity in cultures of the NOX⁺ strain grown under aerobic conditions without nisin induction was similar to that of strain MG1363 (about 0.22 μmol · min⁻¹ · mg of protein⁻¹). Induction with 5 μg of nisin per liter resulted in an 80-fold increase in this enzymatic activity (17 μmol · min⁻¹ · mg of protein⁻¹). Figure 2 shows a selection of ¹³C spectra acquired during the metabolism of 20 mM [6-¹³C]glucose by a cell suspension of the NOX⁺ strain under an oxygen-saturated atmosphere. The concentrations of the end products acetoin, acetate, and lactate were monitored as a function of time along with the concentrations of the intracellular metabolites [1-¹³C]Fru(1,6)P₂, [6-¹³C]Fru(1,6)P₂, G6P, 3-PGA, and PEP during the metabolism of glucose (20 mM) by cell suspensions of the NOX⁺ strain under an oxygen-saturated atmosphere (Fig. 3A and C). Glucose metabolism was rerouted towards the production of acetoin (10.4 ± 0.3 mM) and acetate (12.1 ± 0.7 mM), as well as trace amounts (0.3 ± 0.1 mM) of diacetyl (as determined by ¹³C-NMR analysis of NMR sample extracts). This metabolic shift represented a 60-fold increase in the flux through α-acetolactate synthase (from less than 0.01 in strain MG1363 to 0.60 in the NOX⁺ strain) and a 2-fold increase in the flux through pyruvate dehydrogenase; on the other hand, lactate (2.3 ± 0.5 mM) accounted for only 6% of the carbon in glucose. Once glucose was added, the G6P and Fru(1,6)P₂ concentrations increased to 7 ± 1 and 29 ± 1 mM (averages of four experiments), respectively; the Fru(1,6)P₂ concentration was significantly lower than the maximal Fru(1,6)P₂ concentration reached in strain MG1363 under the same conditions (Fig. 1C). At the onset of glucose exhaustion, the 3-PGA and PEP concentrations increased, leveling off at 36 ± 3 and 15 ± 3 mM, respectively.

In separate experiments, the intracellular P_i and nucleoside triphosphate (NTP) levels, as well as the evolution of the intracellular pH, were monitored in vivo by ³¹P-NMR (Fig. 3E). The concentration of NTP increased from undetectable levels to a maximum of 8.8 ± 0.5 mM and then dropped to undetectable levels soon after glucose exhaustion (at ∼19 min). Concomitant with NTP depletion, a sudden rise in the intracellular P_i concentration from 5 mM to about 22 mM was observed, followed by a gradual increase to the initial level (about 38 mM). Addition of glucose caused a sudden increase in the intracellular pH (1.08 pH units), which also returned slowly to its initial value once glucose was depleted.

In the absence of oxygen, the glucose metabolism of the NOX⁺ strain resembled that of strain MG1363, with 93% of the carbon being converted to lactate (Fig. 3B). The profiles of intracellular metabolites were also identical, although a slightly higher PEP concentration (6 ± 2 mM) was measured in the NOX⁺ strain. The total NADH oxidase activity measured in cell extracts of the NOX⁺ strain grown under anaerobic conditions was 16.5 μmol · min⁻¹ · mg of protein⁻¹, which was 235-fold higher than the total NADH oxidase activity found for strain MG1363 (0.07 μmol · min⁻¹ · mg of protein⁻¹) grown under the same conditions.

Characterization of glucose metabolism in a strain deficient in NADH oxidase.

Glucose metabolism in a strain carrying a deletion of the lactococcal nox-2 gene was also examined by ¹³C-NMR in vivo (Fig. 4). A residual total NADH oxidase activity of 0.025 μmol · min⁻¹ · mg of protein⁻¹ was determined in the cells, probably due to H₂O₂-forming activity (NADH dependent) coupled with peroxidase activity. Also, a contribution by H₂O-forming NADH oxidases other than that encoded by the nox-2 gene cannot be ruled out. Under anaerobic conditions, this strain produced lactate (37.2 ± 1.2 mM) and trace amounts of acetate, ethanol, 2,3-butanediol, and formate during metabolism of 20 mM glucose (Fig. 4). Under aerobic conditions, the amount of acetate produced was significantly higher (3.7 ± 0.6 mM), corresponding to 9% of the carbon flux. The Fru(1,6)P₂ concentration reached maximal levels of 45.5 ± 0.7 mM (anaerobic conditions) and 38.7 ± 0.7 mM (aerobic conditions). In the presence of O₂, the pattern of Fru(1,6)P₂ consumption was similar to that exhibited by strain MG1363, but under anaerobic conditions the decline in the Fru(1,6)P₂ concentration was considerably slower, and the levels of 3-PGA and PEP were below the detection limits of the in vivo NMR procedure. For the corresponding NMR sample extracts, a value of 3 mM was calculated for the intracellular concentration of 3-PGA, but PEP was not detected. Under an oxygen atmosphere the 3-PGA concentration reached a maximum of 25 ± 3 mM; PEP could not be detected in vivo, but a concentration of 3.9 ± 0.3 mM was observed in vitro. These results show that the presence of oxygen had a significant impact on intermediates and end products of glucose metabolism, even in cells with very low total NADH oxidase activity. Table 2 summarizes the data for the maximal values determined for intracellular metabolites in the three strains examined under aerobic and anaerobic conditions. Data for the glucose consumption rate and total NADH oxidase activity are also presented.

The crucial role of pyridine nucleotides in the metabolism of L. lactis means that they must be studied in a noninvasive manner. The evolution of NAD⁺ and NADH concentrations in the NOX⁻ strain during the consumption of glucose (80 mM) under anaerobic conditions was monitored in vivo with a time resolution of 2.2 min (Fig. 4C). Before glucose addition (starved cells) the level of NAD⁺ was 3.3 mM, and it remained constant while glucose was available. At the onset of glucose depletion the NAD⁺ concentration dropped abruptly to 1.2 mM concomitant with a steep decline in the Fru(1,6)P₂ concentration. On the other hand, the level of NADH was below the detection limit of the in vivo NMR technique (0.5 mM) and increased to circa 1.8 mM once glucose was exhausted. After this, the levels of Fru(1,6)P₂ (8 mM), NAD⁺ (1.2 mM), and NADH (1.8 mM) remained unchanged (Fig. 4C).

Effect of pH on the pattern of intracellular metabolites and end products.

A comparison of the end products resulting from the metabolism of glucose by the MG1363 and NOX⁺ strains at different pH values (pH 6.5, 5.5, and 4.5) under aerobic conditions is shown in Table 3; the corresponding data for maximal levels of phosphorylated metabolites are shown in Table 4. The glucose metabolism of strain MG1363 at pH 5.5 and 6.5 was also investigated under anaerobic conditions; no significant pH-dependent changes in the pattern of end products were found (data not shown), although decreased maximal levels of Fru(1,6)P₂ (38 mM), 3-PGA (6 mM), and PEP (<0.5 mM) were observed at pH 5.5.

In strain MG1363, the concentration of acetoin increased from 0.13 to 2.65 mM when the pH was decreased from 6.5 to 5.5, but diacetyl production was not detected. This flavor compound, however, was produced by the NOX⁺ strain, especially at the lower pH values. In strain MG1363, the stimulation of acetoin production at a lower pH was accompanied by a reduction in lactate production, whereas the most pronounced effect in the NOX⁺ strain was the decrease in acetate production. Therefore, lowering the pH to 5.5 under aerobic conditions influenced the distribution of the pyruvate flux via LDH and α-acetolactate synthase in strain MG1363. In contrast, the pH-induced decrease in acetate production by the NOX⁺ strain reflected the pH dependence of α-acetolactate synthase, which becomes more competitive than pyruvate dehydrogenase at lower pH values (41). The lower NADH oxidase activity in the NOX⁺ strain at pH values below 6.5 is likely to be due to inactivation of the enzyme by diacetyl (23).

A remarkable reduction in the intracellular metabolite concentrations was observed in strain MG1363 when the pH was decreased from 6.5 to 5.5 (Table 4); for example, the Fru(1,6)P₂ and 3-PGA concentrations decreased approximately twofold, and the reduction in the PEP concentration was even greater (threefold). Interestingly, in the NOX⁺ strain, the Fru(1,6)P₂ concentration was affected only slightly, but the extents of reduction in the 3-PGA and PEP concentrations were as great as the extents of reduction for strain MG1363.

NADH oxidase affinity for oxygen.

To explain the metabolic behavior of the NOX⁺ strain in the presence of oxygen, it was deemed important to know the kinetic parameters of the overproduced enzyme, H₂O-forming NADH oxidase, for its substrates. The K_m of S. mutans NADH oxidase for NADH was reported to be approximately 25 μM (18), but the affinity for oxygen is unknown. Therefore, the enzyme was purified, and the affinity for oxygen was determined in this study. A K_m of 2 μM for O₂ was found. To our knowledge, this is the first report of the oxygen affinity of any NADH oxidase.