Metabolic Analysis of Glutamate Production by Corynebacterium glutamicum

Metabolic Engineering 1, 224231 (1999) Article ID mben.1999.0122, available online at http:www.idealibrary.com on

Metabolic Analysis of Glutamate Production by Corynebacterium glutamicum Pierre Gourdon and Nicholas D. Lindley 1 Centre de Bioingenierie Gilbert Durand, Centre National de la Recherche ScientifiqueUnite Mixte de Recherche 5504, UR 792 INRA, Institut National des Sciences Appliquees, Complexe Scientifique de Rangueil, F-31077 Toulouse Cedex 4, France Received March 22, 1999; accepted June 22, 1999

The dynamic behavior of the metabolism of Corynebacterium glutamicum during l-glutamic acid fermentation, was evaluated by quantitative analysis of the evolution of intracellular metabolites and key enzyme concentrations. Glutamate production was induced by an increase of the temperature and a final concentration of 80 g l was attained. During the production phase, various other compounds, notably lactate, trehalose, and DHA were secreted to the medium. Intracellular metabolites analysis showed important variations of glycolytic intermediates and NADH, NAD coenzymes levels throughout the production phase. Two phenomena occur during the production phase which potentially provoke a decrease in the glutamate yield: Both the intracellular concentrations of glycolytic intermediates and the NADH NAD ratio increase significantly during the period in which the overall metabolic rates decline. This correlates with the decrease in glutamate yield due in part to the production of lactate and also to the period of the fermentation in which growth no longer occurred.  1999 Academic Press

INTRODUCTION Corynebacterium glutamicum is widely used for the production of amino acids, such as l-glutamic acid and l-lysine. Because of the economic importance of these fermentations, a great deal of effort has been made, in the last decade, to improve our working knowledge of the physiology of this bacterium. Investigations have focused on the regulation of amino acid biosynthetic pathways and the carbon flux distribution within the central metabolism. Various groups have examined the fate of glucose in the central pathways, using NMR (Rollin et al., 1995; Marx et al., 1996; Dominguez et al., 1998), enzymatic (Cocaign-Bousquet et al., 1996) and mathematical modelling (Vallino and Stephanopoulos, 1993; Pons et al., 1996; Park et al., 1997a) 1 To whom correspondence should be addressed at centre de Bioingenierie Gilbert Durand, INSA, Complexe Scientifique de Rangueil, 31077 Toulouse cedex 4, France. Fax: (33) 5 61 55 94 00. E-mail: lindleyinsa-tlse.fr.

1096-717699 30.00 Copyright  1999 by Academic Press All rights of reproduction in any form reserved.

224

approaches. Particular attention have been paid to the pyruvatePEP branching point in an attempt to define the relative contribution of each anaplerotic enzyme for TCA cycle during growth and amino acid synthesis (Park et al., 1997b; Peters-Wendisch et al., 1996). Much of this work has involved either growth of the organism on various substrates or l-lysine production. The metabolic constraints associated with l-glutamic acid production have not been as thoroughly characterized. However, NMR experiments (Ishino et al., 1991; Sonntag et al., 1995; Marx et al., 1997), have reported significant modifications of metabolic flux between exponential growth and l-glutamate production. Most important are a diminished flux through the pentose pathway and an increased flux through isocitrate dehydrogenase during the phase of l-glutamate production. The smaller contribution of the pentose pathway was assumed to occur because of the diminished requirement of NADPH in glutamic acid formation, postulating a correlation between the carbon flux distribution and the energetic demand though this has proved difficult to exploit (Marx et al., 1999). However, these experiments though elegant in their laboratory conception often used either poor production strains or fermentation strategies somewhat different from those suitable for industrial use. Little attempt was made to see how flux distribution varied throughout the production phase, notably at glutamate concentrations of industrial interest. In this study, a temperature-triggered process of glutamate overproduction has been used and dynamic variations in intracellular metabolite pools have been correlated to whole cell kinetic behavior. Important variations of metabolites levels were measured during the fermentation indicating that the production phase cannot be assumed to be a period of quasi-steady-state behavior. This dynamic analysis suggests that strain improvement strategies based upon data obtained during the early phase of production may fail to perform up to expectations due to the modified metabolic constraints intervening later in the fermentation.

Metabolic Engineering 1, 224231 (1999) Article ID mben.1999.0122

Metabolic Constraints during Glutamate Fermentation

MATERIALS AND METHODS Organism and Cultivation The strain used in this work was Corynebacterium glutamicum 2262. Cultivation was done in a 3.5-liter reactor (Chemap). The pH was maintained at 7.6 by automatic addition of NH 4 (12 N). During the production phase, the culture was pulsed with a concentrated glucose solution (500 gl) to avoid periods of glucose limitation. The medium was based on MCGC medium (Von der Osten et al., 1989) although citrate was replaced by deferoxamine. The medium contained glucose (60 gl), Na 2HPO 4 (3 gl), KH 2PO 4 (6 gl), NaCl (2 gl), (NH 4 ) 2SO 4 (8 gl), MgSO 4, 7H 2O (0.4 gl), FeSO 4, 7H 2O (40 mgl), FeCl 3 (4 mgl), ZnSO4, 7H 2O (1 mgl), CuCl 2, 2H 2O (0.4 mgl), MnSO 4, H 2O (2 mgl), (NH 4 ) 6Mo 7O 24, 4 H 2O (0.2 mgl), Na 2B 4O 7 , 10 H 2O (0.4 mgl), CaCl 2 (84 mgl), biotin (2 mgl), thiamine (20 mgl), desferoxamine (3 mgl), and betaine (2 gl). The onset of glutamate production was induced by increasing the growth temperature from 33 to 39.5%C over a 5 min period when the biomass concentration reached 5.6 gl. Analytical Methods Biomass was estimated by absorbance at 650 nm and by a direct gravimetric method following drying of washed cells to constant weight under partial vacuum at 60%C for at least 24 h. In this manner, potential errors in biomass estimation linked to morphological changes were avoided. Sugars and organic acids were determined on the culture supernatant by HPLC using a HPX87H column (Biorad) maintained at 48%C with H 2SO 4 (5 mM) as eluant. Detection was performed by UV and refractometer detectors. Gas phase composition was determined by gas chromatography with a Porapak Q column maintained at 40%C with helium as the carrier gas and catharometer detection. Glutamate and other amino acid concentrations was determined with an AminoQuant 1090 high-pressure liquid chromatography (HewlettPackard) after derivatization by orthophthaladehyde in the presence of 3-mercaptopropionic acid, separation with a C 18 column, and spectrophotometric detection at 338 nm. Enzymatic Assays Most enzyme activities were assayed by spectrophotometric measurement of variation in NAD(P)H concentration at 340 nm (==6.223 M &1 cm &1 ). One unit of activity was the amount of enzyme required to produce 1 nmol of product per min. All enzyme activity were measured in crude cell-free extracts obtained by sonication. Crude 225

extracts were prepared as follows: cell were harvested, washed twice with 0.20 KCl, and resuspended in Tristricarballylate buffer (250 mM, pH 7.8) containing glycerol (300), and MgCl 2 (5 mM). Cells were disrupted by sonication and cell debris was removed by centrifugation at 10,000g, for 10 min at 4%C. The supernatant was used for enzyme assay and the protein concentration of the extract was determined by the Lowry method with bovine serum albumin as standard. Glucose 6-phosphate dehydrogenase was assayed by a method based on that of Sugimoto and Shiio (1987a) in a reaction mixture containing Tris-HCl buffer (100 mM, pH 7.8), MgCl 2 (10 mM), NADP + (0.5 mM), and glucose 6-phosphate (2 mM) as the substrate. 6-phosphogluconate dehydrogenase was assayed by a method based on that of Sugimoto and Shiio (1987b) using the same reaction mixture as above, except that 6-Phosphogluconate (1 mM) was added as the substrate instead of glucose 6-phosphate. Phosphoenolpyruvate carboxylase was assayed by a method based on that of Mori and Shiio (1984) in a reaction mixture containing Tris-HCl buffer (100 mM, pH 7.8), MnSO 4 (5 mM), KHCO 3 (10 mM), NADH 2 (0.15 mM), acetyl coenzyme-A (0.1 mM), 10 +g } ml &1 malate dehydrogenase, and PEP (2 mM). The reaction was started by the addition of PEP. Glyceraldehyde-3-phosphate dehydrogenase was assayed according to the method of Crow and Wittenberger (1979) with the following reaction mixture: NAD (1 mM), sodium arsenate (5 mM), cysteine HCl (5 mM), triethanolamine buffer (125 mM, pH 7.8), dl-glyceraldehyde-3-P (4 mM), and diluted enzyme extract. Malic enzyme was assayed by the method described by Mori and Shiio (1987) with an optimized reaction mixture consisting of phosphate buffer (100 mM, pH 7.8), MgCl 2(5 mM), NADP + (0.6 mM), and malate (40 mM) as the substrate. Assays were initiated by the addition of malate. Isocitrate dehydrogenase was assayed with the following reaction mixture : Tris-HCl buffer (100 mM, pH 7.8), MnSO 4 (5 mM), NADP + (0,6 mM), and isocitrate (10 mM). Assays were initiated by the addition of isocitrate. Glutamate dehydrogenase was assayed using a reaction mixture containing Tris-HCl buffer (100 mM, pH 7.8), NH 4Cl (40 mM), NADPH 2 (0.6 mM), and :-ketoglutarate (10 mM). Nonspecific NADH oxidation was assayed using a reaction mixture containing Tris-HCl buffer (100 mM, pH 7.8), MnSO 4 (5 mM), and NADH 2 (0.3 mM). Extraction and Estimation of Intracellular Metabolite Concentrations The extraction procedure optimized by Dominguez et al. (1998) was used in which cell samples (10 ml) of known cell

Metabolic Engineering 1, 224231 (1999) Article ID mben.1999.0122

Gourdon and Lindley

dry weight, were removed directly from the culture, frozen immediately in liquid nitrogen, and stored at &80%C. After rapid thawing, either HCl (3N) or KOH (5N) were added to give a final pH of 1.5 or 12.5, respectively. The acid extraction procedure was used for all metabolites except NADH and was achieved by incubating the HCl-treated culture (pH 1.5) at 25%C for 10 min before neutralizing to pH 6.8 with KOH (10 N) while agitating vigorously to avoid transiently alkaline pH conditions. After centrifugation at 12,000g (8 min) to remove cell debris (no visible chemical precipitation occurred) the supernatant was used for assays. Extraction procedures were tested with aqueous solutions of all metabolites and loss during extraction was shown to be less than 50 in all cases, though this loss increased if more acidic or prolonged extraction procedures were used. Acid-labile coenzyme NADH were extracted by incubating the KOH-treated culture at pH 12.5 at 25%C for 10 min. After centrifugation at 4%C for 5 min at 12,000g the supernatant was immediately tested for NADH without neutralizing to avoid destruction of NADH by locally high concentrations of acid. Intracellular metabolites were assayed by measuring the increase (or decrease) of NAD(P)H fluorescence, using a spectrofluorometer (Hitachi F 2000). The excitation wavelength was 350 nm and the emission wavelength was 460 nm. The enzyme-coupled assay procedures used were those described by Le Bloas et al. (1993). All metabolite concentrations were given relative to dry cell weight since the significant changes in culture medium composition made estimations of cell volume via the silicon oil centrifugation technique prone to error due to poor pellet formation. RESULTS Fermentation Kinetics After a growth phase of approximately 5 h, glutamate production is induced by a rapid increase of the fermentation temperature (Fig. 1). Throughout the production phase, the temperature was maintained at 39.5\0.5%C. In less than 2 h, the growth rate decreased from 0.58 h &1 to 0.2 h &1 and the specific rate of glutamate production reached a maximum value of 0.55 g } g &1 } h &1 (Fig. 2). Glutamate production and glucose consumption remained maximal for 6 h while the growth rate continued to decrease slowly. During this phase, trehalose, lactate, and dihydroxyacetone were detected at low levels in the medium. The presence of these metabolites indicates metabolic leakage at three levels within glycolysis (glucose 6-phosphate, pyruvate, and triose phosphates). Such overflow metabolism was not due to oxygen limitation since dissolved oxygen was maintained above 200 saturation. After 12 h of fermentation, the specific 226

FIG. 1. Fermentation time course for glutamate overproduction by C. glutamicum. Data points represent biomass (Q), glucose consumed (M), production of glutamate (g), trehalose (G), lactate (m), and DHA (q).

rate of glutamate production decreased progressively in parallel with the specific rate of glucose consumption, coinciding with the stopping of biomass proliferation and an acceleration of lactate production which reached a value of 0.032 gg } h &1. In the final hours of the fermentation, the specific rate of lactate production further increased to 0.06 gg } h &1. Trehalose was produced continuously throughout the production phase and constituted the main coproduct throughout the fermentation. The specific production rate increase regularly from 0.04 gg } h &1 to reach 0.06 gg } h &1 after 14 h and remained constant thereafter. Production of dihydroxyacetone was considerably less important and was maintained throughout the fermentation. Stoichiometric Analysis of the Fermentation Stoichiometric analysis based upon product yields, calculated from data given above, illustrates that metabolic changes occurred throughout the fermentation (Fig. 3). Throughout the fermentation carbon balances were good, i.e., more than 950 of the sugar consumed was accounted for in the major products (biomass, CO 2 , glutamate, trehalose, lactate and dihydroxyacetone). The remainder

Metabolic Engineering 1, 224231 (1999) Article ID mben.1999.0122

Metabolic Constraints during Glutamate Fermentation

FIG. 2. Evolution of specific rates throughout the fermentation. (A) Glucose consumption ( - - ), glutamate production (   ), and growth (www). (B) trehalose production ( - ), lactate production (  ), and DHA production (-------).

was in part associated with a number of minor products present at trace amounts (succinate, :-ketoglutarate, acetate, alanine, glutamine, and N-acetyl-glutamine). The post-temperature shift part of the fermentation can be divided into three distinct phases. The first phase which lasted approximately 2 h is characterized by a rapid shift from growth to glutamate production. A second phase is then initiated until growth stops after 15 h of fermentation. This phase is marked by a reduction of the biomass yield associated with co-metabolite production. It can be speculated that the decrease of the anabolic carbon flux leaving glycolysis for biomass synthesis provokes an increase of the glycolytic flux and probably leads to overflow metabolism. The glutamate yield remains stable during this phase. A final phase begins after the growth arrest with an increase in the yields of lactate and trehalose at the expense of glutamate production. This tendency accelerates in the final 3 h of the fermentation in which a marked decrease in 227

FIG. 3. Instantaneous yields relative to glucose consumption throughout the fermentation. The vertical division lines indicate distinct phases during the glutamate production phase as mentioned in the text. (A) Glutamate (   ), biomass (www), and CO 2 ( - - ). (B) trehalose ( -), lactate (  ), and DHA (-------).

carbon conversion efficiency is seen for glutamate with concomitant increases in the molar yields for lactate, trehalose, and CO 2 . Enzyme Activity Profiles The specific activities of various enzymes, representative of different pathways of central metabolism, were assayed throughout the culture (Fig. 4). A transient decrease of all activities was observed immediately after the temperature shift, corresponding, to an increased cellular protein content, probably due to the induced synthesis of stress proteins. Glutamate dehydrogenase activity remained constant throughout the fermentation. The enzyme concentration seems not to be affected by the shift to glutamate production. A similar profile was obtained when glutamate production was induced by biotin limitation, addition of penicillin G or surfactant (Kawahara et al., 1997). The GAP

Metabolic Engineering 1, 224231 (1999) Article ID mben.1999.0122

Gourdon and Lindley

The two anaplerotic enzymes assayed display a different evolution. The specific activity of malic enzyme decreased rapidly after the temperature shift whereas the PEP carboxylase activity increased to a maximal value between 10 and 15 h of fermentation, when the growth rate is very low. This finding corroborates previous observations concerning the growth rate dependent activity of each anaplerotic enzyme in glucose-grown cells (Cocaign-Bousquet et al., 1996). In view of the rapid disappearance of malic enzyme activity it can probably be concluded that this enzyme plays no major role in TCA cycle replenishment during the greater part of the glutamate production phase. The PEP carboxylase activity is significantly lower than the anaplerotic flux required for glutamate except during the terminal phase of the fermentation once growth had stopped. As concluded elsewhere (Park et al., 1997b; PetersWendisch et al., 1996) a pyruvate carboxylating enzyme is probably ensuring the majority of the anaplerotic flux during the production phase though we were unable to measure reliably this activity in the strain used in the present study. Finally an NADH oxidizing activity, initially believed to be due to NADH oxidase activity (Cocaign-Bousquet et al., 1996) but now believed to be attributable to NADH dehydrogenase components of the respiratory chain (Hertz, 1998) decreased rapidly in parallel to the growth rate. Metabolite Concentrations FIG. 4. Evolution of specific enzyme activities of C. glutamicum throughout the fermentation. (A) Glyceraldehyde 3-phosphate dehydrogenase (s), isocitrate dehydrogenase (S), and glutamate dehydrogenase (Q). (B) 6-phosphoguconate dehydrogenase (m), glucose 6-phosphate dehydrogenase (M), phosphoenolpyruvate carboxylase (G), malic enzyme (g), and NADH oxidizing activity (q). Each value represents the average of three independent determinations with error bars.

dehydrogenase activity increased rapidly during the first half of the production phase prior to stabilising at a value approximately twofold higher than during exponential growth when growth no longer occurred. There is an apparent contradiction between the increase of GAP dehydrogenase specific activity and DHA production indicative of a possible flux limitation through this enzyme though clearly this will depend upon the in vivo regulation of this enzyme activity. The specific activity of isocitrate dehydrogenase increased progressively throughout the production phase. The two dehydrogenation reactions (G6P dehydrogenase and 6PG dehydrogenase) constituting the initial entry into the pentose pathway showed different variations in specific activity profiles: 6PG dehydrogenase behaved similarly to GAP dehydrogenase while G6P dehydrogenase activity decreased markedly once growth ceased. 228

The intracellular concentrations of key phosphorylated glycolytic intermediates, pyruvate and NAD(H) were followed during the fermentation. The concentration of all glycolytic metabolites increased during the second half of the production phase indicating a progressive limitation of the flux through glycolysis (Fig. 5). Both glucose 6-phosphate and fructose 6-phosphate concentrations remained constant for about 6 h after the temperature shift at levels comparable with these measured during exponential growth phase, but increased sharply thereafter throughout the following 12 h until the end of the fermentation to concentrations more than fourfold higher than seen in exponential growth. Fructose 1,6-diphosphate concentration increases sharply during the initial phase of glutamate production and then more slowly throughout the remainder of the production phase. Triose-Ps (GAP and DHAP) levels increase linearly from the onset of the production phase to stabilize transiently at a value of 20 +molg coinciding with the peak of DHA production. Thereafter, the pool concentration increased sharply to reach a maximal value of 80 +molg when growth ceased: value maintained throughout the remainder of the fermentation. Phosphorylated intermediates situated downstream of GAP but upstream of PEP were at all times lower than the analytical precision

Metabolic Engineering 1, 224231 (1999) Article ID mben.1999.0122

Metabolic Constraints during Glutamate Fermentation

FIG. 5. Concentrations of intracellular metabolites in cell samples taken throughout the fermentation. Each data point represents the average of four independent determinations with error bars.

threshold (<0.2 +mol). The manner in which the pool of PEP evolved throughout the production phase was similar to that observed for sugar phosphates (glucose 6-phosphate and fructose 6-phosphate) possibly involving the type of coordinated regulation of phosphofructokinase and pyruvate kinase activities described for Lactobacillus bulgericus (Le Bras et al., 1998). The pyruvate pool shows 229

a progressive but small increase in concentration throughout the production phase until the final hours of the fermentation during which the pyruvate concentration increases rapidly, concomitant with the increased rate of lactate production. Although the intracellular concentration of NAD remained approximately constant throughout the fermentation

Metabolic Engineering 1, 224231 (1999) Article ID mben.1999.0122

Gourdon and Lindley

the NADH concentration increased in a pattern similar to that observed for many of the glycolytic intermediates, i.e., a slight increase throughout the initial period of the production phase with a more marked increase in the later period of glutamate production. This leads to a significant modification of the NADHNAD ratio, known to play a major role in the in vivo regulation of glycolysis and notably as an inhibitor of pyruvate and GAP dehydrogenases (Snoep et al., 1992; Garrigues et al., 1997; Dominguez et al., 1998) while such an increase in intracellular NADH concentration is also known to activate lactate dehydrogenase activity (Garrigues et al., 1997) which is constitutively expressed in C. glutamicum but increases progressively throughout the production phase to reach activities threefold higher than those measured in exponentially growing cells (results not shown). It would appear likely that this increase in NADHNAD ratio, probably due to a decreased respiratory capacity since NADH oxidation decreases rapidly after the temperature shift, provokes many of the intracellular modifications of metabolite pools and the consequences this has on whole cell behavior. DISCUSSION Metabolic engineering strategies attempt to obtain a realistic view of the metabolic constraints associated with the specific operating conditions used in industrial fermentations to pragmatically identify targets for rational improvement of either the process or the biological component of the process. However, this approach frequently assumes that metabolite overproduction obeys a rigid steady-state type behavior and that rate-controlling reactions remain so over the entire production phase despite the frequently observed changes in kinetic behavior. Fermentation processes are intrinsically changing systems, or at least in the more commonly used batch systems, in which the physicochemical environment seen by the cells is being constantly modified, not least by the accumulation of desired metabolites. During the fermentation, the temperatureinduced overflow of glutamate leads to an important accumulation of products (notably glutamic acid) with associated osmotic stress. This progressive increase in osmolarity is associated with trehalose production whose intracellular accumulation is known to contribute to osmoprotective mechanisms (Guillouet and Engasser, 1995) but is frequently reported to be excreted into the medium during amino acid overproduction using C. glutamicum (Ishino et al., 1991; Park et al., 1997a). In the temperature induced production phase it appears that the capacity to maintain cellular division is important for process performance. The progressive loss of cell division and modified catabolic potential is seen in the slowing down 230

of the specific metabolic rates and even more clearly in the shift from glutamate production to alternative products in the final phase of the fermentation. This is particularly noticeable for lactate production which accompanies the increased NADHNAD ratio. Whole cell kinetics and intracellular analysis of metabolites and enzyme activities are similar to what would be expected under oxygen limiting conditions though only trace amounts of succinate, indicative of anoxic conditions (Dominguez et al., 1993), were measured during the production phase. Indeed the situation appears to be more closely akin to the behavior of C. glutamicum during growth on fructose in which both DHA and lactate accumulate due to the modified intracellular flux partition and the resulting increase in glycolysis (Dominguez et al., 1998). This was attributed to an increased NADHNAD ratio and the control exerted over dehydrogenation reactions such that glycolysis becomes saturated. In this case, the overflow of lactate and DHA is again correlated to an increased NADHNAD ratio though in this case this would appear to be associated to the diminished capacity to oxidize NADH. Future work needs to establish to what extent respiratory activity is affected under such conditions. Clearly increasing product yields in such a fermentation will probably depend upon our capacity to apply correctly the metabolic knowledge that can be gained in physiological studies. Analysis of the initial period of glutamate production would no doubt lead to credible targets for improving carbon conversion efficiency, but the loss of metabolic activity observed throughout the second half of the fermentation would most probably attenuate the anticipated gains. An alternative approach, derived directly from the metabolic analysis used in this study would be to modify the process constraints so as to maintain a higher level of metabolic activity and thereby prolong the period of high rate and high yield of glutamate production, retarding the increased NADHNAD ratio and associated consequences for product formation and sugar consumption rate. ACKNOWLEDGEMENTS The authors thank Mrs. Monique Suderie for valuable technical assistance. This work was made possible by financial assistance from ORSAN-Amylum France, the CNRS, and as part of the work package supported by the European Union Grant BIO4-CT96-01145.

REFERENCES Cocaign-Bousquet, M., Guyonvarch, A., and Lindley, N. D. (1996). Growth rate dependent modulation of carbon flux through central metabolism and the kinetic consequences for glucose-limited chemostat cultures of Corynebacterium glutamicum. Appl. Env. Microbiol. 62, 429436.

Metabolic Engineering 1, 224231 (1999) Article ID mben.1999.0122

Metabolic Constraints during Glutamate Fermentation

Crow, W. L., and Wittenberg, C. L. (1979). Separation and properties of NAD +- and NADP +-dependent glyceraldehyde-3-phosphate dehydrogenase from Streptococcus mutans. J. Biol. Chem. 254, 11341142. Dominguez, H., Nezondet, C., Lindley, N. D., and Cocaign, M. (1993). Modified carbon flux during oxygen-limited growth of Corynebacterium glutamicum and the consequences for amino acid production. Biotechnol. Lett. 15, 449454. Dominguez, H., Rollin, C., Guyonvarch, A., Guerquin-Kern, J. L., Cocaign-Bousquet, M., and Lindley, N. D. (1998). Carbon-flux distribution in the central metabolic pathways of Corynebacterium glutamicum during growth on fructose. Eur. J. Biochem. 254, 96102. Garrigues, C., Loubiere, P., Lindley, N. D., and Cocaign-Bousquet, M. (1997). Control of the shift from homolactic acid to mixed-acid fermentation in Lactococcus lactis : Predominant role of the NADHNAD + ratio. J. Bacteriol. 179, 52825287. Guillouet, S., and Engasser, J. M. (1995). Growth of Corynebacterium glutamicum in glucose-limited continuous cultures under high osmotic pressure. Influence of growth rate on the intracellular accumulation of proline, glutamate and trehalose. Appl. Microbiol. Biotechnol. 44, 496500. Hertz, P. F. (1998). ``Cha@^ ne respiratoire et deshydrogenases flaviniques chez Corynebacterium melassecola ATCC 17965: Contribution a l'etude des voies d'utilisation de coenzymes intervenant dans les reactions d'oxydoreduction.'' Ph.D. thesis. Universite Paris Sud, France. Ishino, S., Shimomura-Nishimuta, J., Yamaguchi, K., Shirahata, K., and Araki, K. (1991). 13C Nuclear magnetic resonance studies of glucose metabolism in L-glutamic acid and L-lysine fermentation by Corynebacterium glutamicum. J. Gen. Appl. Microbiol. 37, 157165. Kawahara, Y., Takahashi-Fuke, K., Shimizu, E., Nakamatsu, T., and Nakamori, S. (1997). Relationship between the glutamate production and the activity of 2-oxoglutarate dehydrogenase in Brevibacterium lactofermentum. Biosci. Biotech. Biochem. 61, 11091112. Lebloas, P., Guilbert, N., Loubiere, P., and Lindley, N. D. (1993). Growth inhibition and pyruvate overflow during glucose metabolism of Eubacterium limosum are related to a limited capacity to reassimilate CO 2 by the acetyl-CoA pathway. J. Gen. Microbiol. 139, 18611868. Le Bras, G., de la Torre, F., Branny, P., and Garel, J. R. (1998). Enzymatic and genetic regulation of glycolysis in Lactobacillus delbrueckii subsp. bulgaricus. Lait 78, 8590. Marx, A., de Graaf, A. A., Wiechert, W., Eggeling, L., and Sahm, H. (1996). Determination of the fluxes in the central metabolism of Corynebacterium glutamicum by nuclear magnetic resonance spectroscopy combined with metabolic balancing. Biotechnol. Bioeng. 49, 111129. Marx, A., Striegel, K., de Graaf, A. A., Sahm, H., and Eggeling, L. (1997). Response of the central metabolism of Corynebacterium glutamicum to different flux burdens. Biotechnol. Bioeng. 56, 168180.

231

Marx, A., Eikmanns, B. J., Sahm, H., de Graaf, A. A., and Eggeling, L. (1999). Response of the central metabolism in Corynebacterium glutamicum to the use of an NADH-dependent glutamate dehydrogenase. Metabolic Engineering 1, 3548. Mori, M., and Shiio, I. (1984). Production of aspartic acid and enzymatic alteration in pyruvate kinase mutants of Brevibacterium flavum. Agric. Biol. Chem. 48, 11891197. Mori, M., and Shiio, I. (1987). Pyruvate formation and sugar metabolism in an amino acid-producing bacterium, Brevibacterium flavum. Agric. Biol. Chem. 51, 129138. Park, S. M., Sinskey, A. J., and Stephanopoulos, G. (1997a). Metabolic and physiological studies of Corynebacterium glutamicum mutants. Biotechnol. Bioeng. 55, 864879. Park, S. M., Shaw-Reid, C., Sinskey, A. J., and Stephanopoulos, G. (1997b). Elucidation of anaplerotic pathways in Corynebacterium glutamicum via 13C-NMR spectroscopy and GC-MS. Appl. Microbiol. Biotechnol. 47, 430440. Peters-Wendisch, P. G., Wendisch, V. F., de Graaf, A. A., Eikmanns, B. J., and Sahm, H. (1996). C3-carboxylation as an anaplerotic reaction in phosphoenolpyruvate-deficient Corynebacterium glutamicum. Arch. Microbiol. 165, 387396. Pons, A., Dussap, C. G., Pequignot, C., and Gros, J. B. (1996). Metabolic flux distribution in Corynebacterium glutamicum ATCC 17965 for various carbon sources. Biotechnol. Bioeng. 51, 177189. Rollin, C., Morgant, V., Guyonvarch, A., and Guerquin-Kern, J. L. (1995). 13 C-NMR studies of Corynebacterium melassecola metabolic pathways. Eur. J. Biochem. 227, 488493. Snoep, J. L., Teixeira de Mattos, M. J., Starrenburg, M., and Hugenholtz, J. (1992). Isolation, characterisation and physiological role of the pyruvate dehydrogenase complex and :-acetolactate synthase of Lactococcus lactis subsp. diacetylactis. J. Bacteriol. 174, 48314841. Sonntag, K., Schwinde, J., de Graaf, A. A., Marx, A., Eikmanns, B. J., Wiechert, W., and Sahm, H. (1995). 13C NMR studies of the fluxes in the central metabolism of Corynebacterium glutamicum during growth and overproduction of amino acids in batch cultures. Appl. Env. Microbiol. 44, 489495. Sugimoto, S., and Shiio, I. (1987a). Regulation of glucose 6-phosphate dehydrogenase in brevibacterium flavum. Agric. Biol. Chem. 51, 101108. Sugimoto, S., and Shiio, I. (1987b). Regulation of 6-phosphogluconate dehydrogenase in brevibacterium flavum. Agric. Biol. Chem. 51, 12571263. Vallino, J., and Stephanopoulos, G. (1993). Metabolic flux distributions in Corynebacterium glutamicum during growth and lysine overproduction. Biotechnol. Bioeng. 41, 633646. Von der Osten, C. H., Gioannetti, C., and Sinskey, A. J. (1989). Design of a defined medium for growth of Corynebacterium glutamicum in which citrate facilitates iron uptake. Biotechnol. Lett. 11, 1116.