INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

In the baking industry, frozen-dough technology has recently been accepted due to its advantages, which include supplying oven-fresh bakery products to consumers and improving labor conditions for bakers (15). Ordinary commercial baker's yeast is generally susceptible to damage during frozen storage and does not retain sufficient leavening ability after frozen storage. Prefermented frozen doughs yield low-quality bread due to the poor gassing power of freeze-injured yeast. Several types of freeze-tolerant yeasts have been found in natural sources (10, 14). Although freeze-tolerant yeasts have also been obtained by conventional mutation procedures (20, 22, 32), bread baked with freeze-tolerant yeast strains obtained by conventional mutation procedures has less taste and flavor than bread baked with the parent strains. In this study our goal was to construct freeze-tolerant baker's yeast strains from commercial strains by using DNA recombinant techniques without degrading the other beneficial properties of the yeasts (e.g., taste and flavor).

We focused on trehalose as a cryoprotectant for yeast cells. The reason for this is that in the yeast Saccharomyces cerevisiae the disaccharide trehalose is thought to be a critical determinant of stress tolerance, including freezing dehydration, and heat shock tolerance, because its concentration increases under certain adverse environmental conditions (7, 9, 13, 29). The cellular levels of trehalose are controlled by a balance between the trehalose-synthesizing and trehalose-hydrolyzing enzymes (23). The bifunctional enzyme called trehalose-6-phosphate synthetase/trehalose-6-phosphate phosphatase synthesizes trehalose in the cytosol by condensing glucose-6-phosphate and UDP-glucose (23). The following two enzymes are capable of hydrolyzing trehalose: a neutral cytosolic trehalase (designated Nth1p) and an acidic trehalase (designated Ath1p) (19). Both Nth1p and Ath1p have been purified from S. cerevisiae, and the corresponding genes have been cloned and sequenced (1, 8, 17). Maximum Nth1p activity occurs at pH 6.0 to 7.0, and this activity is regulated by the RAS/adenylate cyclase signal transduction pathway, which converts the inactive form to the phosphorylated, active form (19). Nth1p plays a role in protecting cells against heat shock (25). In contrast to cytosolic Nth1p, Ath1p is a vacuolar protein whose optimum pH is 4.5 (21); this protein is necessary for the phenotype of growth on trehalose (i.e., trehalose utilization) (26). Biswas and Ghosh reported that Ath1p is present in trehalase-sucrase aggregates (2). However, the regulatory pathway that controls Ath1p activity is not known.

In order to determine the effect of trehalase disruption during the baking process, we constructed trehalase-deficient diploid strains derived from commercial baker's yeast strains. Because commercial strains differ from laboratory strains in many properties, such as trehalose content, stress tolerance, leavening ability, and flavor formation, commercial strains were used in this work. In fact, neither haploid nor diploid strains derived from laboratory strains can be used for baking tests due to their poor gassing powers. We found that trehalase-deficient mutants accumulate higher levels of trehalose than their parent strains accumulate under the culture conditions that are optimal for high trehalose contents and that trehalose accumulation correlates with high freeze tolerance in frozen-dough baking. We also found that it may be possible to use mutant yeast strains obtained by gene disruption for commercial applications.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains, media, growth conditions, and mating. The strains of baker's yeast used in this study are listed in Table 1. Yeast extract-peptone-dextrose (YPD) medium contained (per liter) 10 g of yeast extract (Difco Laboratories), 20 g of peptone (Difco), and 20 g of glucose. YPD agar contained 2% Bacto Agar (Difco) in addition to the components mentioned above. Synthetic dextrose medium contained (per liter) 1.7 g of yeast nitrogen base without amino acids and ammonium sulfate (Difco), 5 g of ammonium sulfate, and 20 g of glucose. Liquid fermentation (LF) medium contained (per liter) 100 g of sucrose. For negative selection of ura3 mutants, we used 5-fluoroorotic acid (5-FOA) plates containing (per liter) 1 g of 5-FOA, 6.7 g of yeast nitrogen base without amino acids (Difco), 20 g of glucose, 120 mg of uracil, and 20 g of Bacto Agar (Difco). Cane molasses medium contained (per liter) 30 g of sugar (calculated as sucrose), 5.66 g of urea, and 0.62 g of NaH₂PO₄.

Plasmid construction and disruption of trehalase genes. We constructed the trehalase-deficient strains by using one-step gene disruption, which involved double recombination events at the homologous site (28). Transformation was carried out by using the LiCl methods described by Schiestl and Gietz (31). A fragment used for the ATH1 disruption procedure was constructed as follows. The ATH1 open reading frame (ORF) was amplified by PCR from strain T7 chromosomal DNA by using primers designed on the basis of the ATH1 sequences described by Destruelle et al. (8). The sequence of primer 1 was 5'-CATCCACTGGGAGTGGTTTC-3', and the sequence of primer 2 was 5'-CGAGATGATTGCCAATGTCT-3'. The PCR product was then cloned into pGEM-T (Promega), which yielded pCI1. Next, pCI1 was digested with HindIII with only one site of the ATH1 ORF in the plasmid. The URA3 gene isolated as a HindIII fragment from YEp24 was inserted into the HindIII site of pCI1, which yielded pCY1. pCY1 was linearized with PvuII prior to transformation. A fragment used for the NTH1 disruption procedure was constructed as follows. NTH1 was obtained from the CEN fragment of chromosome III included in YCp50 by using a gene eviction method (33). A KpnI-EcoRI 770-bp fragment of NTH1 was inserted into the KpnI-EcoRI site of pUC19, which yielded pNTH1-KE. The URA3 gene isolated as a HindIII fragment from YEp24 prior to blunting with a Klenow fragment was ligated to pNTH1-KE, which had been blunted with a Klenow fragment after it was digested with XhoI. The resulting plasmid, pNTHd1, was linearized with EcoRI and KpnI prior to transformation into the parent strain.

DNA isolation, Southern blot analysis, and molecular biology methods. Yeast DNA was isolated essentially as described by Hereford et al. (12). Southern hybridization was carried out by using a Hybond-N nylon membrane (Amersham) and an ECL direct nucleic acid labeling and detection system (Amersham) according to the manufacturer's instructions. Standard molecular biology techniques were adapted as described by Sambrook et al. (30).

Assays for trehalose, protein, and trehalase activity. Yeast cells were collected by centrifugation (1,000 × g) for 10 min and were washed three times with cold (4°C) distilled water. After 0.5 M trichloroacetic acid was used to extract the trehalose from the yeast cells, the amount of trehalose was measured by the anthrone method (4). A crude lysate of yeast cells was prepared by disruption with glass beads. The protein concentration of the crude lysate was determined by the Bradford method (3). The activities of neutral trehalase and acid trehalase were assayed by using crude extracts of cells as described by Mittenbuhler and Holzer (21).

Dehydration of yeast cells. First, yeast cells were dehydrated to 68% relative humidity, and then a sorbitan monostearate emulsion was added at a proportion equivalent to 1.5% of the yeast solids content (11). The preparation obtained was extruded through a screen having a mesh size of 0.4 mm. The yeast was dried to a solids content of 95% by fluidization in a current of hot, dry air. Dehydration was carried out within 50 min; during this process the temperature of the yeast did not exceed 38°C.

Amount of gas produced in dough. To determine leavening ability, the volume (in milliliters) of carbon dioxide gas produced after incubation for 2 h at 30°C was measured by using a Formograph AF-1000 apparatus (Atto Co., Tokyo, Japan). The volume was normalized to standard conditions (30°C, 101.3 kPa). The low-sugar dough formula contained 100 parts of bread-making flour (13.5% moisture basis), 6 parts of sugar, 2 parts of NaCl, 2 parts of yeast (67% moisture basis), and 65 parts of water. The ingredients were mixed for 2 min with a Swanson type mixer (Eberhardt-Denver Co., Denver, Colo.), and the resulting dough was divided into 40-g pieces, which were kept in polyethylene bags. The dough pieces were fermented for 60 min at 30°C before they were frozen at 20°C. The frozen dough was thawed for 30 min at 30°C before gas production was measured.

Baking. White bread and sweet bread were prepared by the straight dough method. The formula for the white bread was as follows: 100 parts of bread-making flour (13.5% moisture basis), 2 parts of NaCl, 2 parts of yeast cells (67% moisture basis), 5 parts of shortening, and 66 parts of water. The ingredients were mixed for 2 min with a Swanson type pin mixer (Eberhardt-Denver Co.). After fermentation for 115 min at 30°C, the dough was punched twice and molded (relative humidity, 75%). The dough was divided, and a sample which was used for a freeze tolerance test was stored for 1 week at 20°C. The frozen dough was thawed for 30 min at 30°C before final proofing. After the dough was proofed for 55 min at 38°C, it was baked for 25 min at 200°C. The formula for the sweet bread was as follows: 70 parts of bread-making flour (13.5% moisture basis), 30 parts of semihard flour (13.5% moisture basis), 25 parts of sucrose, 0.7 parts of NaCl, 4 parts of yeast cells (67% moisture basis), 6 parts of shortening, 2 parts of nonfat dry milk powder, and 54 parts of water. After the dough was proofed for 45 min at 38°C, it was baked for 20 min at 200°C. The loaf volume (in milliliters) and weight (in grams) were measured within 1 h after baking and the specific volume (in milliliters per gram of bread) was calculated. Both the external characteristics and the internal characteristics of the bread were evaluated by using the method of the Japan Yeast Industry Association.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Construction of trehalase mutants derived from commercial baker's yeast strains by gene disruption. Table 1 shows the strains used in this study. The haploid strains which we used for trehalase gene disruption and to form diploid strains were selected from commercial baker's yeast strains based on high levels of transformation efficiency and good fermentation properties. The diploid strains that were obtained by mating the haploid strains had gassing powers equivalent to the gassing powers of commercial baker's yeast strains (data not shown). Since the parent strains were prototrophs, the spontaneous ura3 mutants were obtained by 5-FOA negative selection (5). The ura3 mutants derived from four haploid strains (two mating type a strains and two mating type  strains) were used for transformation. The nth1 mutants and the ath1 mutants were obtained by transformation of the haploid strains with fragments containing NTH1 or ATH1. Southern blot analysis confirmed that gene disruption occurred. Genomic DNAs isolated from transformants and parent strains were digested with EcoRI and were hybridized by using either an NTH1 probe or an ATH1 probe (data not shown). The nth1 ath1 double mutants were constructed by transformation of nth1 ura3 mutants, which were obtained by spontaneous 5-FOA reselection from nth1 strains by using the fragment for ATH1 disruption. The Southern blot analysis confirmed that the double mutation was present (Fig. 1).

View larger version (18K): [in this window] [in a new window]   FIG. 1.   Southern blot analysis of nth1 ath1 double mutant. Total yeast genomic DNA was digested with EcoRI, electrophoresed on 1% agarose gels, transferred to a nylon membrane, and hybridized with a 1.4-kb EcoRI-KpnI fragment containing the NTH1 ORF (A) and a 1.1-kb fragment containing the ATH1 ORF prepared by PCR in which primer 1 and primer 2 (B) were used as probes. Lane 1, T118; lane 2, AT418; lane 3, T128; lane 4, AT428.

Intracellular accumulation and degradation of trehalose in the trehalase mutants. In this study, cells of trehalase mutants and parent strains were grown in continuously fed batch cultures in order to obtain a high trehalose content, which simulated the industrial yeast production process. At various times during growth, samples were removed, and the intracellular trehalose levels were measured (see above). We tested two different genetic backgrounds, T118 and T128. In the T118 background, the nth1 (T154), ath1 (A318), and nth1 ath1 (AT418) strains accumulated substantially more trehalose than parent strain T118 accumulated (Fig. 2A). At the end of the stationary phase, the amounts of trehalose accumulated by these mutants were about twice the amounts accumulated by the parent strains. Under the experimental conditions used, all of the mutant strains exhibited nearly identical trehalose accumulation characteristics. Contrary to expectations, the NTH1 ATH1 double disruption had no detectable synergistic effect on trehalose accumulation. In contrast, the amount of trehalose that accumulated was slightly increased by the trehalase disruption of T128, which accumulated more trehalose than T118 accumulated (Fig. 2B). To determine the effect of trehalase disruption on the baking of frozen dough, we used the T118 background mutant series as a model for an analysis of stress tolerance in dough.

View larger version (18K): [in this window] [in a new window]   FIG. 2.   Intracellular trehalose accumulation by trehalase mutants of T118 (A) and trehalase mutants of T128 (B). Samples were obtained at various growth phases, including the seed culture (SC), the early logarithmic phase (EL), the late logarithmic phase (LL), the early stationary phase (ES), and the late stationary phase (LS). (A) Symbols: , T118; , T154; , A318; , AT418. (B) Symbols: , T128; , T164; , A328; , AT428. The results are means ± standard deviations based on three independent experiments.

View larger version (20K): [in this window] [in a new window]   FIG. 3.   Changes in intracellular trehalose content during fermentation in LF medium at 30°C. Symbols: , T118; , T154; , A318; , AT418. The results are means ± standard deviations based on three independent experiments.

Stress tolerance of the trehalase mutants. Intracellular accumulation of trehalose is believed to increase a yeast's tolerance to freezing. We assessed this tolerance in white bread doughs at various frozen-storage times by measuring the rates of CO₂ production in thawed doughs that contained trehalase mutants and parent strains (Fig. 4). All of the trehalase mutants retained more gassing power than the parent strains retained. Among the trehalase mutants, the nth1 strain had the highest freeze tolerance associated with a high intracellular trehalose level during fermentation and produced 100 ml of CO₂ even after 3 weeks frozen storage.

View larger version (18K): [in this window] [in a new window]   FIG. 4.   Gassing power in frozen doughs stored for different periods of time at 20°C. The doughs were thawed, and then CO2 production at 30°C was measured for 2 h. Symbols: , T118; , T154; , A318; , AT418. The results are means ± standard deviations based on three independent experiments.

View larger version (14K): [in this window] [in a new window]   FIG. 5.   Gassing power after dehydration to a relative humidity of 5% and rehydration. CO2 production by each yeast in LF medium at 30°C was measured for 2 h. The results are means ± standard deviations based on three independent experiments.

Baking with trehalase mutants obtained by gene disruption. The baking properties of trehalase mutants were assessed by measuring the specific volumes of baked white bread and sweet bread. The specific volume represented the leavening property of baker's yeast because it increased in proportion to the residential gassing power in the dough. White breads prepared with either the parent strains or the trehalase mutants had similar specific volumes, but the sweet breads prepared with the trehalase mutants had higher specific volumes than the sweet breads prepared with the parent strains (Table 3). In general, the fermentation ability of baker's yeast is inhibited in doughs containing osmolytes, such as sugar and salt, at high concentrations. In fact, the specific volume of sweet bread prepared with a parent strain was significantly less than the specific volume of white bread prepared with the parent strain (Table 3). However, the specific volumes of sweet breads prepared with the trehalase mutants remained high. A high level of trehalose may protect the ability to ferment from high osmolarity.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In the baking industry, frozen-dough methods require yeast strains that have a high freeze tolerance and also retain desirable properties, such as flavor and strong gassing power. Although freeze-tolerant yeasts have been obtained by mutation (20, 22, 32), breads prepared with freeze-tolerant yeast strains obtained by conventional mutation procedures generally lack adequate taste and flavor. In this study, we examined the possibility that mutant yeast strains obtained by gene disruption could be used commercially.

The yeast strains available for frozen dough accumulate high levels of trehalose (9, 13, 29). Several studies have shown that there is a close correlation between trehalose levels and tolerance to freezing (16, 24). Trehalose is believed to be a stress-related metabolite and may have a stress protection function (18, 23). We attempted to regulate cellular levels of trehalose by avoiding hydrolyzing enzymes. We constructed, for the first time, diploid homozygous trehalase mutants derived from commercial baker's yeast strains by molecular biological techniques. During fermentation, the trehalase mutations suppressed degradation of intracellular trehalose and significantly improved freeze tolerance (Fig. 3 and 4). Parent strain T118 had good leavening and flavor formation abilities, but it had a freeze tolerance that was slightly poorer than the freeze tolerance of commercial baker's yeast available from the market for frozen-dough baking. Strain T154, which is an NTH1 disruptant derived from T118, was developed to have a freeze tolerance level similar to that of commercial frozen-dough baking yeast, as well as the other beneficial properties of T118. White bread dough containing T154 can be frozen for up to 2 weeks without a loss of gassing power. Our results suggest that the ath1 and nth1 mutations are effective for breeding of baker's yeast. Kim et al. have reported that ATH1 disruption of laboratory strains is more effective for trehalose accumulation than NTH1 disruption and that the ATH1 NTH1 double disruption has a synergistic effect (16). In contrast, we did not observe either of these effects. One possible reason for this is that the effects of trehalase disruption on trehalose accumulation may depend on the genetic background of the strains themselves. For the trehalase mutants that we studied, the patterns of trehalose accumulation varied from strain to strain, depending on the parent used (Fig. 2). Our results show that the effect of trehalase disruption on trehalose accumulation was less for strains that have high-trehalose backgrounds.

Kim et al. have reported that ATH1 disruption results in tolerance to dehydration, as well as tolerance to freezing (16). In this study, the nth1 and ath1 mutants exhibited significantly increased dry tolerance compared with the parent strain, but the nth1 ath1 strain exhibited significantly decreased dry tolerance. One possible explanation for why the double disruptant exhibited lower dry tolerance is that trehalase activity may be a requirement for rehydration.