was genetically engineered to synthesize a membrane-bound nicotinamide nucleotide transhydrogenase that catalyzes the proton-coupled transfer of reducing equivalents between the NAD(H) and NADP(H) coenzyme systems. Production of the recombinant transhydrogenase was confirmed by a Western blot analysis (Fig. ). Membranes isolated from the recombinant strains exhibited reduction of 3-acetylpyridine-NAD+
by NADPH and by NADH in the presence of NADP+
, which is consistent with the normal catalytic function of the recombinant protein (Fig. ). Purified bovine transhydrogenase also catalyzes reduction of 3-acetylpyridine-NAD+
by NADH in the presence of NADPH (14
), which was interpreted as involving a cyclic reduction-oxidation cycle of bound NADP(H) (14
). A similar reaction has been found for the E. coli
enzyme at low pH values (18
). This pH-dependent catalytic mechanism has been observed for all known membrane-bound transhydrogenases.
Increased formation of 2-oxoglutarate was observed in strain TN24 expressing pnt
genes at a high level, compared to reference strains TN1 and TN3 (30
) and strain TN25, which express only one copy of the pnt
genes (Table ). The absence of increased formation of 2-oxoglutarate in strain TN25 indicated that the compound was synthesized by strain TN24 due to the high transhydrogenase activity. When ammonium is the nitrogen source, this compound and 2-oxoglutarate are converted into glutamate by glutamate dehydrogenase during oxidation of NADPH to NADP+
). A high rate of conversion of NADPH and NAD+
and NADH in strain TN24 by the transhydrogenase decreases the intracellular pool of NADPH and is expected to result in a reduced rate for the reaction catalyzed by the NADPH-dependent glutamate dehydrogenase. If the rate of synthesis of 2-oxoglutarate is not affected by the change in the intracellular NADPH concentration, the reduction in consumption of 2-oxoglutarate by glutamate dehydrogenase results in its secretion. Hence, secretion of 2-oxoglutarate from strain TN24 indicated that NADPH was consumed at a high rate by a transhydrogenase in this strain. The low level of expression of transhydrogenase in strain TN25 did not result in high enough consumption of NADPH to result in secretion of appreciable amounts of 2-oxoglutarate.
Introduction of the high-copy-number plasmid YEpPGKαTDHβ into strain TN3 resulted in a decrease in the maximal specific growth rate, which indicated that the level of expression of transhydrogenase affects the growth rate. Since the rate of conversion of 2-oxoglutarate to glutamate was reduced in strain TN24, the reduced maximal specific growth rate could have been due to the decrease in glutamate synthesis necessary for biomass synthesis.
There was an increase in the glycerol yield, from 0.093 C-mol/C-mol of glucose in strains TN1 and TN3 (30
) to 0.110 C-mol/C-mol of glucose in the low-expression strain TN25 and 0.118 C-mol/C-mol of glucose in the high-expression strain TN24 (Table ) (C-mol means n;umber of gram atoms of carbon in the aoumt of compound in question). Glycerol is formed during anaerobic growth of wild-type S. cerevisiae
, so that excess NADH formed during the synthesis of biomass and organic acids can be redoxidized. In strains TN25 and TN24 the reaction catalyzed by transhydrogenase represents a new pathway for NADH formation since the enzyme catalyzed the reaction in the direction from consumption of NADPH and NAD+
towards formation of NADP+
and NADH. In strains TN24 and TN25 this resulted in the observed increase in the glycerol yield. The reduced biomass yield of strain TN24 reduced the net formation of NADH, but the effect on the glycerol yield was not quantified.
The acetate yield in strain TN24 was greater than the acetate yield in strains TN1 and TN3 (30
). In the last two steps of acetate synthesis, pyruvate is converted into acetaldehyde and then into acetate by pyruvate decarboxylase and the NADP+
-dependent cytoplasmic aldehyde dehydrogenase, respectively, so that 1 mol of NADPH is synthesized per mol of acetate formed. The greater acetate formation in strain TN24 may reflect a regulatory mechanism that compensates for the consumption of NADPH by the transhydrogenase. A similar effect has been observed in recombinant S. cerevisiae
strains that express XYL1
, which encodes an NADPH-consuming xylose reductase (26
In the strain with a high level of expression of the pnt genes, the NADPH/NADP+ ratio decreased from 5.0 to 2.0, which supported the hypothesis that the transhydrogenase converted NADPH into NADH in strain TN24 (Table ). The values also indicated that the presence of transhydrogenase did not result in equilibrium. The increased consumption of NADPH did not lead to an increased level of NADP+, so [NADPH] plus [NADP+] decreased by a factor of two in strain TN24. This change indicates that there was strict regulation of the NADP+ concentration in the cell. Furthermore, the concentration of NAD+ increased in the strain expressing the transhydrogenase, despite increased formation of NADH. This change may have been due to very rigid regulation of the NADH/NAD+ ratio, as indicated by the constant value for this ratio in the strains.
Expression of a transhydrogenase influenced the rates of formation of glycerol and acetate and the rate of consumption of 2-oxoglutarate. These reactions occurred in the cytoplasm. Thus, the changes in the flux of glycerol and acetate and in the rate of consumption of 2-oxoglutarate must have been due to changes in the rates of production of the nucleotides in the cytoplasm. This suggests that nucleotide binding sites of the membrane-bound transhydrogenase are located in this compartment.
The transhydrogenase expressed in S. cerevisiae
converted NADPH and NAD+
and NADH, indicating that the reverse (leftward) reaction of equation 1 occurred. This reaction direction suggests that the Δp across the ER, where most of the transhydrogenase is located in S. cerevisiae
, is insufficient to drive the transhydrogenase forward reaction (equation 1). There is no method to directly measure the ER lumenal pH. The results of indirect measurements and predictions of the ER microenvironment based on characteristics of several ER proteins, however, suggest that the ER pH is approximately 7 (22
). To our knowledge, the membrane potential across the ER membrane has not been determined, which precludes an estimate of Δp.
We do not know why the recombinant transhydrogenase tends to accumulate in rough ER and is not delivered to the plasma or vacuolar membranes. Misfolded or unassembled proteins tend to accumulate in the ER and to degrade rapidly (16
), but our work indicates that the recombinant protein has an intact catalytic function. A certain sequence necessary for proper assembly for exit from rough ER could be missing in the sequence of E. coli
). Protein movement from the thin phospholipid-rich ER Golgi membranes to the thick sterol- and sphingolipid-rich plasma membranes could be limited by the length of the transmembrane domains (29
). It is possible that by adding sorting signals present in integral membrane proteins from yeast and by using screens to identify genes encoding proteins that facilitate the sorting events the membrane-bound transhydrogenase could be directed to the plasma membrane or to the vacuolar membrane. Particularly in the plasma membrane, the Δp may be sufficient to drive equation 1 in the desired direction. Assuming that the measured nucleotide levels are representative of the cytoplasm and that n
in equation 1 is 1, the required proton gradient would be on the order of 100 mV. Also, the kinetics would be favored by a large proton gradient.
We found that a functional membrane transhydrogenase can be synthesized in S. cerevisiae. However, this protein is not delivered to the plasma membrane but seems to accumulate in internal membrane systems, mainly in the rough ER (Fig. ). Our results show that product formation by S. cerevisiae expressing a transhydrogenase from E. coli is affected. We found that the yields of glycerol and acetic acid increased and the yield of ethanol decreased, indicating that a reversed reaction (equation 1) catalyzed by the transhydrogenase occurred. The intracellular concentrations of the four nucleotides confirmed that the degree of reduction of the NADP(H) pool is higher than the degree of reduction of the NAD(H) pool in S. cerevisiae expressing the transhydrogenase and that the Δp at the location of the recombinant transhydrogenase probably is insufficient to make a forward reaction (equation 1) possible.