In the fermentation of lignocellulose hydrolysates to ethanol, two major problems are encountered: the fermentation of the pentose sugar xylose, and the presence of microbial inhibitors. Xylose can be directly fermented with yeasts, such as Pachysolen tannophilus, Candida shehatae, and Pichia stipis, or by isomerization of xylose to xylulose with the enzyme glucose (xylose) isomerase (XI; EC 5.3.1.5), and subsequent fermentation with bakers' yeast, Saccharomyces cerevisiae. The direct fermentation requires low, carefully controlled oxygenation, as well as the removal of inhibitors. Also, the xylose-fermenting yeasts have a limited ethanol tolerance. The combined isomerization and fermentation with XI and S. cerevisiae gives yields and productivities comparable to those obtained in hexose fermentations without oxygenation and removal of inhibitors. However, the enzyme is not very stable in a lignocellulose hydrolysate, and S. cerevisiae has a poorly developed pentose phosphate shunt. Different strategies involving strain adaptation, and protein and genetic engineering adopted to overcome these different obstacles, are discussed.

Glucose and xylulose fermentation and product formation by Saccharomyces cerevisiae were compared in batch culture under anaerobic conditions. In both cases the main product was ethanol, with glycerol, xylitol, and arabitol produced as by-products. During glucose and xylulose fermentation, 0.74 and 0.37 g of cell mass liter-', respectively, were formed. In glucose-fermenting cells, the carbon balance could be closed, whereas in xylulose-fermenting cells, about 25% of the consumed sugar carbon could not be accounted for. The rate of sugar consumption was 3.94 mmol g of initial biomass-' h-1 for glucose and 0.39 mmol g of initial biomass-' h-' for xylulose. Concentrations of the intermediary metabolites fructose-1,6-diphosphate (FDP), pyruvate (PYR), sedoheptulose 7-phosphate (S7P), erytrose 4-phosphate, citrate (CIT), fumarate, and malate were compared for both types of cells. Levels of FDP, PYR, and CIT were lower, and levels of S7P were higher in xylulose-fermenting cells. After normalization to the carbon consumption rate, the levels of FDP were approximately the same, whereas there was a significant accumulation of S7P, PYR, CIT, and malate, especially of S7P, in xylulose-fermenting cells compared with in glucose-fermenting cells. In the presence of 15 ,iM iodoacetate, an inhibitor of the enzyme glyceraldehyde-3-phosphate dehydrogenase (EC 1.2.1.12), FDP levels increased and S7P levels decreased in xylulose-assimilating cells compared with in the absence of the inhibitor, whereas fermentation was slightly slowed down. The specific activity of transaldolase (EC 2.2.1.2), the pentose phosphate pathway enzyme reacting with S7P and glyceraldehyde-3-phosphate, was essentially the same for both glucoseand xylulose-fermenting cells. It was, however, several orders of magnitude lower than that reported for a Torula yeast and Candida utilis. The presence of iodoacetate did not influence the activity of transaldolase in xylulose-fermenting cells. The results are discussed in terms of a competition between the pentose phosphate pathway and glycolysis for the common metabolite, glyceraldehyde-3-phosphate, which would explain the low rates of xylulose assimilation and ethanol production from xylulose by S. cerevisiae. * Corresponding author. iodoacetate (IA), an inhibitor of glyceraldehyde-3-phosphate (G3P) dehydrogenase (EC 1.2.1.12) (4, 6, 36), on metabolite levels and the specific activity of the PPP enzyme transaldolase (EC 2.2.1.2) were studied in order to elucidate the importance of the pool of G3P on the rate of xylulose fermentation in S. cerevisiae. MATERIALS AND METHODS Organism. S. cerevisiae ATCC 24860 was maintained at 4°C on slants containing yeast extract (3 g liter-1; Difco Laboratories, Detroit, Mich.), Bacto-Peptone (5 g liter-1; Difco), agar (20 g liter-1; Difco), malt extract (3 g liter'; Difco), and glucose or xylose (10 g liter-').

In vitro metabolism of D-xylulose and D-glucose in extracts obtained from D-glucoseand D-xylulosefermenting Saccharomyces cerevisiae cells was investigated with 10and 100-fold-increased activity of the enzyme transaldolase (EC 2.2.1.2). The rate of sugar consumption was the same in most cases, whereas the rate of ethanol formation decreased with increased levels of transaldolase. The formation of glycerol, pentitols, and acetic acid was not dependent on added transaldolase but was dependent on the sugar used as the growth substrate and on the sugar used in the in vitro metabolism experiments. The carbon balance showed that the dissimilated carbon could not be accounted for in products when transaldolase was added. The concentration of D-fructose-1,6.-diphosphate in the extracts was not influenced by added transaldolase but was higher with D-xylulose than with D-glucose. Levels of pyruvate, comparable with the two substrates, decreased with increasing levels of transaldolase. Exogenously added transaldolase decreased D-sedoheptulose-7-phosphate levels when D-xylulose was the substrate. The results are discussed in relation to the dissimilation of carbon through the upper part of glycolysis and the pentose phosphate pathway.

Concentrations of the intracellular intermediary metabolites fructose 1,6-diphosphate, pyruvate, citrate, and malate in free and calcium alginate-immobilized cells of Saccharornyces cerevisiae fermenting D-glucose anaerobically were determined when the sugar up-take rate and the ethanol production rate were constant No cell growth was observed and the fermentation yields and fermentation rates were the same in both types of cells. The concentrations of intermediary intracellular metabolites were also identical for the two types of fermenting cells.