Growth of Corynebacterium glutamicum on fructose was significantly less than that obtained on glucose, despite similar rates of substrate uptake. This was in part due to the production of overflow metabolites (dihydroxyacetone and lactate) but also to the increased production of CO 2 during growth on fructose. These differences in carbon-metabolite accumulation are indicative of a different pattern of carbon-flux distribution through the central metabolic pathways. Growth on glucose has been previously shown to involve a high flux (Ͼ 50% of total glucose consumption) via the pentose pathway to generate anabolic reducing equivalents. NMR analysis of carbon-isotope distribution patterns of the glutamate pool after growth on 1-13 C-or 6- 13C-enriched fructose indicates that the contribution of the pentose pathway is significantly diminished during exponential growth on fructose with glycolysis being the predominant pathway (80 % of total fructose consumption). The increased flux through glycolysis during growth on fructose is associated with an increased NADH/NAD ϩ ratio susceptible to inhibit both glyceraldehyde-3-phosphate dehydrogenase and pyruvate dehydrogenase, and provoking the overflow of metabolites derived from the substrates of these two enzymes. The biomass yield observed experimentally is higher than can be estimated from the apparent quantity of NADPH associated with the pentose pathway and the flux through isocitrate dehydrogenase, suggesting an additional reaction yielding NADPH. This may involve a modified tricarboxylic acid cycle involving malic enzyme, expressed to significantly higher levels during growth on fructose than on glucose, and a pyruvate carboxylating anaplerotic enzyme.Keywords : Corynebacterium glutamicum; fructose metabolism; NMR analysis; NADH/NAD ϩ ratio; overflow metabolism.For several decades, Corynebacterium glutamicum and re-as to more efficiently supply the specific biosynthetic pathways with the necessary carbon precursors and coenzymes. lated species have been exploited industrially for the production of various amino acids. Improvements of the fermentation strateAlthough many enzymes of intermediary metabolism have been isolated from Brevibacterium flavum and characterised pregies employed and of the bacterial strains by genetic engineering techniques have been achieved, leading to progressively increas-dominantly by the team of Shiio [5Ϫ7], global carbon-flux models based on regulatory or energetic principles have only reing rates of production and/or yields [1]. These strategies have often been based upon overcoming the natural feedback regula-ceived attention recently. Over the last five years, various groups have examined how glucose is catabolised by the central pathtion mechanisms specific to each biosynthetic pathway [2Ϫ4] and have enabled a detailed understanding of these biochemical ways using NMR [8Ϫ10], enzymatic [11] and mathematical modelling [12,13] approaches. A general consensus opinion has sequences to be established. It is now apparent that further impr...

The amino acid producing bacterium Corynebucferirun glufumicum accumulated lactate, succinate and acetate under oxygen-limited growth conditions. Significant restructuring of carbon flux through the central metabolic pathways occurred with a notable decrease in pentose pathway flux and the operation of the TCA cycle in a reductive mode. Simultaneous consumption of residual sugar and organic acids took place when oxygen sufficient conditions were restored though amino acids yields were signiticantly perturbed. lntrodllction For several decades, Corynebacterium glutamicum or related species, have been used industrially for the production of various amino acids. Under certain conditions in which cell membrane permeability is modified via biotine limitation or surfactant addition these bacteria accumulate glutamic acid to high concentrations in the medium broth (Kikuchi and Nakao, 1986). Empirical optimisation of the fermentation strategy has enabled the final concentrations achieved to exceed 100 g/l while traditional genetic techniques have enabled the range of amino acids produced under appropriate fermentation conditions to be extended. Further progress to improve either product yields or specific production rates would now seem to depend on the capacity to modify carbon flux distribution within the central metabolic network to overcome specific limitations. A major limitation to the pragmatic application of this type of metabolic engineering approach is the relative absence of any real data concerning the regulatory mechanisms controlling carbon distribution through the diverse pathways of intermediary metabolism. The stoechiometric modelling techniques used by Vallino & Stephanopoulos (1991) postulated a high carbon flux through the pen&se pathway (5560% of carbon flux from glucose-6-P) during glucose catabolism: figure recently confirmed by NTvIR studies (A Guyonvarch, pers comm). The central metabolism of C.gluramicum is therefore somewhat different to the 'normal metabolism' model, based mostly on E. co/i drtti.in which carbon flux through the pentose pathway is generally believed to be only 1540% of the total G6P available. Such a difference has been attributed to the apparent absence of the NADmADP transhydrogenase enzyme in C. glutamicum and hence the necessity for a direct production of NADPH to satisfy the requirement for anabolic reactions. While such a metabolism is well adapted to the biotechnologists' requirements it must also be born in mind that such modified carbon flux implies a greatly modified control structure and hence the need to establish a number of regulatory details before the rational improvement of the micro-organism can proceed along defined directives. At the moment, much of 449

No abstract