Genetic and physiological evidence for the existence of a second glycolytic pathway in yeast parallel to the phosphofructokinase-aldolase reaction sequence

Abstract

Yeast mutants lacking phosphofructokinase activity because of a defect in one of the two genes PFK1 and PFK2 can still perform glycolysis and produce ethanol. However, they differ from normal wild-type yeast in several ways. After a transfer from a sugar-free to a glucose medium, wild-type cells start to produce ethanol right away, mutants only after a lag period of about 90 min. About two-thirds of the carbon atoms released as CO2 from wild-type cells derive from glucose carbon atoms 3 and 4. Mutants with a single defect in one of the two phosphofructokinase genes PFK1 and PFK2 show no such a preferential contribution of these two C-atoms of glucose. All six C-atoms contribute almost equally to CO2 production. We have isolated mutants that block glycolysis in single pfk1 and pfk2 mutants. They could be located in three different genes called BYP1, BYP2 and BYP3 (BYP for bypass). In a byp1 mutant, CO2 derived almost exclusively from C-atoms 3 and 4 of glucose. This is what the classical concept of yeast glycolysis predicts. During a search for metabolites accumulating in pfk and byp mutants, we found sedoheptulose-7-phosphate, a pentosephosphate cycle intermediate not detectable in wild-type cells. An analysis of enzymes acting in the direct oxidation of glucose-6-phosphate and in the pentosephosphate cycle did not show any defects in those activities. It is hypothesized that the pentosephosphate cycle not only functions, in providing phosphorylated derivatives of tetroses and pentoses for biosynthetic needs, but also plays an important role in sugar catabolism and fermentation. This hypothesis also implies that the reaction sequency catalyzed by phosphofructokinase and aldolase covers only part of the total catabolic flux.