Metabolic Flux Ratio Analysis of Genetic and Environmental Modulations of Escherichia coli Central Carbon Metabolism

The response of Escherichia coli central carbon metabolism to genetic and environmental manipulation has been studied by use of a recently developed methodology for metabolic flux ratio (METAFoR) analysis; this methodology can also directly reveal active metabolic pathways. Generation of fluxome data arrays by use of the METAFoR approach is based on two-dimensional (13)C-(1)H correlation nuclear magnetic resonance spectroscopy with fractionally labeled biomass and, in contrast to metabolic flux analysis, does not require measurements of extracellular substrate and metabolite concentrations. METAFoR analyses of E. coli strains that moderately overexpress phosphofructokinase, pyruvate kinase, pyruvate decarboxylase, or alcohol dehydrogenase revealed that only a few flux ratios change in concert with the overexpression of these enzymes. Disruption of both pyruvate kinase isoenzymes resulted in altered flux ratios for reactions connecting the phosphoenolpyruvate (PEP) and pyruvate pools but did not significantly alter central metabolism. These data indicate remarkable robustness and rigidity in central carbon metabolism in the presence of genetic variation. More significant physiological changes and flux ratio differences were seen in response to altered environmental conditions. For example, in ammonia-limited chemostat cultures, compared to glucose-limited chemostat cultures, a reduced fraction of PEP molecules was derived through at least one transketolase reaction, and there was a higher relative contribution of anaplerotic PEP carboxylation than of the tricarboxylic acid (TCA) cycle for oxaloacetate synthesis. These two parameters also showed significant variation between aerobic and anaerobic batch cultures. Finally, two reactions catalyzed by PEP carboxykinase and malic enzyme were identified by METAFoR analysis; these had previously been considered absent in E. coli cells grown in glucose-containing media. Backward flux from the TCA cycle to glycolysis, as indicated by significant activity of PEP carboxykinase, was found only in glucose-limited chemostat culture, demonstrating that control of this futile cycle activity is relaxed under severe glucose limitation.     

Aerobic glucose metabolism of Saccharomyces kluyveri: growth, metabolite production, and quantification of metabolic fluxes.

The growth and product formation of Saccharomyces kluyveri was characterized in aerobic batch cultivation on glucose. At these conditions it was found that ethyl acetate was a major overflow metabolite in S. kluyveri. During the exponential-growth phase on glucose ethyl acetate was produced at a constant specific rate of 0.12 g ethyl acetate per g dry weight per hour. The aerobic glucose metabolism in S. kluyveri was found to be less fermentative than in S. cerevisiae, as illustrated by the comparably low yield of ethanol on glucose (0.08 +/- 0.02 g/g), and high yield of biomass on glucose (0.29 +/- 0.01 g/g). The glucose metabolism of S. kluyveri was further characterized by the new and powerful techniques of metabolic network analysis. Flux distributions in the central carbon metabolism were estimated for respiro-fermentative growth in aerobic batch cultivation on glucose and respiratory growth in aerobic glucose-limited continuous cultivation. It was found that in S. kluyveri the flux into the pentose phosphate pathway was 18.8 mmole per 100 mmole glucose consumed during respiratory growth in aerobic glucose-limited continuous cultivation. Such a low flux into the pentose phosphate pathway cannot provide the cell with enough NADPH for biomass formation which is why the remaining NADPH will have to be provided by another pathway. During batch cultivation of S. kluyveri the tricarboxylic acid cycle was working as a cycle with a considerable flux, that is in sharp contrast to what has previously been observed in S. cerevisiae at the same growth conditions, where the tricarboxylic acid cycle operates as two branches. This indicates that the respiratory system was not significantly repressed in S. kluyveri during batch cultivation on glucose.     

Engineering of the redox imbalance of Fusarium oxysporum enables anaerobic growth on xylose

Dissimilatory nitrate reduction metabolism, of the natural xylose-fermenting fungus Fusarium oxysporum, was used as a strategy to achieve anaerobic growth and ethanol production from xylose. Beneficial alterations of the redox fluxes and thereby of the xylose metabolism were obtained by taking advantage of the regeneration of the cofactor NAD(+) during the denitrification process. In batch cultivations, nitrate sustained growth under anaerobic conditions (1.21 g L(-1) biomass) and simultaneously a maximum yield of 0.55 moles of ethanol per mole of xylose was achieved, whereas substitution of nitrate with ammonium limited the growth significantly (0.15 g L(-1) biomass). Using nitrate, the maximum acetate yield was 0.21 moles per mole of xylose and no xylitol excretion was observed. Furthermore, the network structure in the central carbon metabolism of F. oxysporum was characterized in steady state. F. oxysporum grew anaerobically on [1-(13)C] labelled glucose and unlabelled xylose in chemostat cultivation with nitrate as nitrogen source. The use of labelled substrate allowed the precise determination of the glucose and xylose contribution to the carbon fluxes in the central metabolism of this poorly described microorganism. It was demonstrated that dissimilatory nitrate reduction allows F. oxysporum to exhibit typical respiratory metabolic behaviour with a highly active TCA cycle and a large demand for NADPH.     

Amino acid biosynthesis and metabolic flux profiling of Pichia pastoris.

Amino acid biosynthesis and central carbon metabolism of Pichia pastoris were studied using biosynthetically directed fractional (13)C labeling. Cells were grown aerobically in a chemostat culture fed at two dilution rates (0.05 h(-1), 0.16 h(-1)) with glycerol as the sole carbon source. For investigation of amino acid biosynthesis and comparison with glycerol cultivations, cells were also grown at 0.16 h(-1) on glucose. Our results show that, firstly, amino acids are synthesized as in Saccharomyces cerevisiae. Secondly, biosynthesis of mitochondrial pyruvate via the malic enzyme is not registered for any of the three cultivations. Thirdly, transfer of oxaloacetate across the mitochondrial membrane appears bidirectional, with a smaller fraction of cytosolic oxaloacetate stemming from the mitochondrial pool at the higher dilution rate of 0.16 h(-1) (for glucose or glycerol cultivation) when compared to the glycerol cultivation at 0.05 h(-1). Fourthly, the fraction of anaplerotic synthesis of oxaloacetate increases from 33% to 48% when increasing the dilution rate for glycerol supply, while 38% is detected when glucose is fed. Finally, the cultivation on glucose also allowed qualitative comparison with the flux ratio profile previously published for Pichia stipitis and S. cerevisiae grown on glucose in a chemostat culture at a dilution rate of 0.1 h(-1). This provided a first indication that regulation of central carbon metabolism in P. pastoris and S. cerevisiae might be more similar to each other than to P. stipitis.     

The metabolic architecture of plant cells. Stability of central metabolism and flexibility of anabolic pathways during the growth cycle of tomato cells

The changes in the intermediary metabolism of plant cells were quantified according to growth conditions at three different stages of the growth cycle of tomato cell suspension. Eighteen fluxes of central metabolism were calculated from (13)C enrichments after near steady-state labeling by a metabolic model similar to that described in Dieuaide-Noubhani et al. (Dieuaide-Noubhani, M., Raffard, G., Canioni, P., Pradet, A., and Raymond, P. (1995) J. Biol. Chem. 270, 13147-13159), and 10 net fluxes were obtained directly from end-product accumulation rates. The absolute flux values of central metabolic pathways gradually slowed down with the decrease of glucose influx into the cells. However, the relative fluxes of glycolysis, the pentose-P pathway, and the tricarboxylic acid cycle remained unchanged during the culture cycle at 70, 28, and 40% of glucose influx, respectively, and the futile cycle of sucrose remained high at about 6-fold the glucose influx, independently from carbon nutritional conditions. This natural resistance to flux alterations is referred to as metabolic stability. The numerous anabolic pathways, including starch synthesis, hexose accumulation, biosynthesis of wall polysaccharides, and amino and organic acid biosynthesis were comparatively low and variable. The phosphoenolpyruvate carboxylase flux decreased 5-fold in absolute terms and 2-fold in relation to the glucose influx rate during the culture cycle. We conclude that anabolic fluxes constitute the flexible part of plant cell metabolism that can fluctuate in relation to cell demands for growth.     

Metabolic engineering under uncertainty--II: analysis of yeast metabolism

Yeast metabolism has been used extensively in scientific investigations and industrial applications. Understanding the properties of the yeast metabolic network is crucial, yet unaccomplished due to its high complexity, the different culture conditions, and the uncertainty associated with kinetic parameters. We recently developed a computational and mathematical framework using Monte Carlo method in which parameter uncertainty can be addressed through large-scale sampling procedure. This framework was applied on the compartmentalized central carbon pathways of Saccharomyces cerevisiae metabolism considering the growth environment of batch and chemostat reactor and integrating information from metabolic flux analysis. Statistical analysis of the results indicates that yeast cells growing in batch culture condition exhibit dramatically different control schemes from those growing in a chemostat. The difference is mainly due to the feedback introduced by the constraints of the chemostat. The control of the enzymes on the rates of the substrate uptake, product excretion, and cell growth and its practical implication are discussed. Clustering of the reaction steps according to the similarity of their responses to enzyme activity perturbations reveals functional coupling of metabolic reactions.     

Application of metabolic flux analysis for the identification of metabolic bottlenecks in the biosynthesis of penicillin-G

A detailed stoichiometric model was developed for growth and penicillin-G production in Penicillium chrysogenum. From an a priori metabolic flux analysis using this model it appeared that penicillin production requires significant changes in fluxes through the primary metabolic pathways. This is brought about by the biosynthesis of carbon precursors for the beta-lactan nucleus and an increased demand for NADPH, mainly for sulfate reduction. As a result, significant changes in flux partitioning occur around four principal nodes in primary metabolism. These are located at: (1) glucose-6-phosphate; (2) 3-phosphoglycerate; (3) mitochondrial pyruvate; and (4) mitochondrial isocitrate. These nodes should be regarded as potential bottlenecks for increased productivity. The flexibility of these principal nodes was investigated by experimental manipulation of the fluxes through the central metabolic pathways using a high-producing strain of P. chrysogenum. Metabolic fluxes were manipulated through growth of the cells on different substrates in carbon-limited chemostat culture. Metabolic flux analysis, based on measured input and output fluxes, was used to calculate the fluxes around the principal nodes. It was found that, for growth on glucose, ethanol, and acetate, the flux partitioning around these nodes differed significantly. However, this had hardly any effect on penicillin productivity, showing that primary carbon metabolism is not likely to contain potential bottlenecks. Further experiments were performed to manipulate the total metabolic demand for the cofactor nicotinamide adenine dinucleotide phosphate (NADPH). NADPH demand was increased stepwise by cultivating the cells on glucose or xylose as the carbon source combined with either ammonia or nitrate as the nitrogen source, which resulted in a stepwise decrease of penicillin production. This clearly shows that, in penicillin fermentation, possible limitations in primary metabolism reside in the supply/regeneration of cofactors (NADPH) rather than in the supply of carbon precursors.     

The protease inhibitor chagasin of Trypanosoma cruzi adopts an immunoglobulin-type fold and may have arisen by horizontal gene transfer

Methylglyoxal metabolism was studied during Saccharomyces cerevisiae grown with D-glucose as the sole carbon and energy source. Using for the first time a specific assay for methylglyoxal in yeast, metabolic fluxes of its formation and D-lactate production were determined. D-Glucose consumption and ethanol production were determined during growth. Metabolic fluxes were also determined in situ, at the glycolytic triose phosphate levels and glyoxalase pathway. Maximum fluxes of ethanol production and glucose consumption correspond to maxima of methylglyoxal and D-lactate formation fluxes during growth. Methylglyoxal formation is quantitatively related to glycolysis, representing 0.3% of the total glycolytic flux in S. cerevisiae.     

Involvement of nitrogen metabolism in the triggering of ethanol fermentation in aerobic chemostat cultures of Saccharomyces cerevisiae.

We have investigated whether central nitrogen metabolism may influence the triggering of ethanol fermentation in Saccharomyces cerevisiae strain CEN.PK122 grown in the presence of different N-sources (ammonia, glutamate, or glutamine) under conditions in which the carbon to nitrogen (C : N) ratio was varied. An exhaustive quantitative evaluation of yeast physiology and metabolic behavior through metabolic flux analysis (MFA) was undertaken. It is shown that ethanol fermentation is triggered at dilution rates, D (growth rate), significantly lower (D=0.070 and 0.074 h(-1) for glutamate and glutamine, respectively, and D=0.109 h(-1) for ammonia) under N- than C-limitation (approximately 0.18 h(-1) for all N-sources). A characteristic specific rate of glucose influx, q(Glc), for each N-source at Dc, i.e., just before the onset of respirofermentative metabolism, was determined (approximately 2.0, 1.5, and 2.5, for ammonia, glutamate, and glutamine, respectively). This q(Glc) was independent of the nutritional limitation though dependent on the nature of the N-source. The onset of fermentation occurs when this "threshold q(Glc)" is overcome. The saturation of respiratory activity appears not to be associated with the onset of fermentation since q(O(2)) continued to increase after Dc. It was remarkable that under respirofermentative conditions in C-limited chemostat cultures, the glucose consumed was almost completely fermented with biomass being synthesized from glutamate through gluconeogenesis. The results obtained show that the enzyme activities involved in central nitrogen metabolism do not appear to participate in the control of the overflow in carbon catabolism, which is driven toward ethanol production. The role of nitrogen metabolism in the onset of ethanol fermentation would rather be realized through its involvement in setting the anabolic fluxes directed to nitrogenous macromolecules. It seems that nitrogen-related anabolic fluxes would determine when the threshold glucose consumption rate is achieved after which ethanol fermentation is triggered.     

Comparative study on central metabolic fluxes of <Emphasis Type="Italic">Bacillus megaterium</Emphasis> strains in continuous culture using <Superscript>13</Superscript>C labelled substrates

Fluxes of central carbon metabolism [glycolysis, pentose phosphate pathway (PPP), tricarboxylic acid cycle (TCA cycle), biomass formation] were determined for several Bacillus megaterium strains (DSM319, WH320, WH323, MS941) in C- and N-limited chemostat cultures by ¹³C labelling experiments. The labelling patterns of proteinogenic amino acids were analysed by GC/MS and therefrom flux ratios at important nodes within the metabolic network could be calculated. On the basis of a stoichiometric metabolic model flux distributions were estimated for the different B. megaterium strains used at various cultivation conditions. Generally all strains exhibited similar metabolic flux distributions, however, several significant changes were found in (1) the glucose flux entering the PPP via the oxidative branch, (2) the reversibilities within the PPP, (3) the relative fluxes of pyruvate and acetyl-CoA fed to the TCA cycle, (4) the fluxes around the pyruvate node involving a futile cycle.     

Changes in the metabolome of Saccharomyces cerevisiae associated with evolution in aerobic glucose-limited chemostats

The effect of culture age on intra- and extracellular metabolite levels as well as on in vitro determined specific activities of enzymes of central carbon metabolism was investigated during evolution for over 90 generations of Saccharomyces cerevisiae CEN.PK 113-7D in an aerobic glucose/ethanol-limited chemostat at a specific dilution rate of 0.052 h(-1). It was found that the fluxes of consumed (O2, glucose/ethanol) and secreted compounds (CO2) did not change significantly during the entire cultivation period. However, morphological changes were observed, leading to an increased cellular surface area. During 90 generations of chemostat growth not only the residual glucose concentration decreased, also the intracellular concentrations of trehalose, glycolytic intermediates, TCA cycle intermediates and amino acids were found to have decreased with a factor 5-10. The only exception was glyoxylate which showed a fivefold increase in concentration. In addition to this the specific activities of most glycolytic enzymes also decreased by a factor 5-10 during long-term cultivation. Exceptions to this were hexokinase, phosphofructokinase, pyruvate kinase and 6-phosphogluconate dehydrogenase of which the activities remained unchanged. Furthermore, the concentrations of the adenylate nucleotides as well as the energy charge of the cells did not change in a significant manner. Surprisingly, the specific activities of glucose-6-phosphate dehydrogenase (G6PDH), malate synthase (MS) and isocitrate lyase (ICL) increased significantly during 90 generations of chemostat cultivation. These changes seem to indicate a pattern where metabolic overcapacities (for reversible reactions) and storage pools (trehalose, high levels of amino acids and excess protein in enzymes) are lost during the evolution period. The driving force is proposed to be a growth advantage in the absence of these metabolic overcapacities.     

Comparative metabolic analysis of lactate for CHO cells in glucose and galactose

t-PA producing CHO cells have been shown to undergo a metabolic shift when the culture medium is supplemented with a mixture of glucose and galactose. This metabolic change is characterized by the reincorporation of lactate and its use as an additional carbon source. The aim of this work is to understand lactate metabolism. To do so, Chinese hamster ovary cells were grown in batch cultures in four different conditions consisting in different combinations of glucose and galactose. In experiments supplemented with glucose, only lactate production was observed. Cultures with glucose and galactose consumed glucose first and produced lactate at the same time, after glucose depletion galactose consumption began and lactate uptake was observed. Comparison of the metabolic state of cells with and without the shift by metabolic flux analysis show that the metabolic fluxes distribution changes mostly in the reactions involving pyruvate metabolism. When not enough pyruvate is being produced for cells to support their energy requirements, lactate dehydrogenase complex changes the direction of the reaction yielding pyruvate to feed the TCA cycle. The slow change from high fluxes during glucose consumption to low fluxes in galactose consumption generates intracellular conditions that allow the influx of lactate. Lactate consumption is possible in cell cultures supplemented with glucose and galactose due to the low rates at which galactose is consumed. Evidence suggests that an excessive production and accumulation of pyruvate during glucose consumption leads to lactate production and accumulation inside the cell. Other internal conditions such as a decrease in internal pH, forces the flow of lactate outside the cell. After metabolic shift the intracellular pool of pyruvate, lactate and H⁺ drops permitting the reversal of the monocarboxylate transporter direction, therefore leading to lactate uptake. Metabolic analysis comparing glucose and galactose consumption indicates that after metabolic shift not enough pyruvate is produced to supply energy metabolism and lactate is used for pyruvate synthesis. In addition, MFA indicates that most carbon consumed during low carbon flux is directed towards maintaining energy metabolism.     

Clostridium cellulolyticum: model organism of mesophilic cellulolytic clostridia

Clostridium cellulolyticum ATCC 35319 is a non-ruminal mesophilic cellulolytic bacterium originally isolated from decayed grass. As with most truly cellulolytic clostridia, C. cellulolyticum possesses an extracellular multi-enzymatic complex, the cellulosome. The catalytic components of the cellulosome release soluble cello-oligosaccharides from cellulose providing the primary carbon substrates to support bacterial growth. As most cellulolytic bacteria, C. cellulolyticum was initially characterised by limited carbon consumption and subsequent limited growth in comparison to other saccharolytic clostridia. The first metabolic studies performed in batch cultures suggested nutrient(s) limitation and/or by-product(s) inhibition as the reasons for this limited growth. In most recent investigations using chemostat cultures, metabolic flux analysis suggests a self-intoxication of bacterial metabolism resulting from an inefficiently regulated carbon flow. The investigation of C. cellulolyticum physiology with cellobiose, as a model of soluble cellodextrin, and with pure cellulose, as a carbon source more closely related to lignocellulosic compounds, strengthen the idea of a bacterium particularly well adapted, and even restricted, to a cellulolytic lifestyle. The metabolic flux analysis from continuous cultures revealed that (i) in comparison to cellobiose, the cellulose hydrolysis by the cellulosome introduces an extra regulation of entering carbon flow resulting in globally lower metabolic fluxes on cellulose than on cellobiose, (ii) the glucose 1-phosphate/glucose 6-phosphate branch point controls the carbon flow directed towards glycolysis and dissipates carbon excess towards the formation of cellodextrins, glycogen and exopolysaccharides, (iii) the pyruvate/acetyl-CoA metabolic node is essential to the regulation of electronic and energetic fluxes. This in-depth analysis of C. cellulolyticum metabolism has permitted the first attempt to engineer metabolically a cellulolytic microorganism.     

13C-Labeled metabolic flux analysis of a fed-batch culture of elutriated Saccharomyces cerevisiae

This study addresses the question of whether observable changes in fluxes in the primary carbon metabolism of Saccharomyces cerevisiae occur between the different phases of the cell division cycle. To detect such changes by metabolic flux analysis, a 13C-labeling experiment was performed with a fed-batch culture inoculated with a partially synchronized cell population obtained through centrifugal elutriation. Such a culture exhibits dynamic changes in the fractions of cells in different cell cycle phases over time. The mass isotopomer distributions of free intracellular metabolites in central carbon metabolism were measured by liquid chromatography-mass spectrometry. For four time points during the culture, these distributions were used to obtain the best estimates for the metabolic fluxes. The obtained flux fits suggested that the optimally fitted split ratio for the pentose phosphate pathway changed by almost a factor of 2 up and down around a value of 0.27 during the experiment. Statistical analysis revealed that some of the fitted flux distributions for different time points were significantly different from each other, indicating that cell cycle-dependent variations in cytosolic metabolic fluxes indeed occurred.