Ability of Saccharomyces cerevisiae MC87-46 to assimilate isomaltose and its effects on sake taste

Recently, wild strains of Saccharomyces cerevisiae isolated from a variety of natural resources have been used to make bread, beer, wine, and sake. In the current study, we isolated wild S. cerevisiae MC strain from the carnation (Dianthus caryophyllus L) flower and produced sake using its cerulenin-resistant mutant strain MC87-46. Then, we characterized the components, including ethanol, amino acids, organic acids, and sugars, in the fermented sake. Sake brewed with MC87-46 is sweet owing to the high content of isomaltose, which was at a concentration of 44.3 mM. The low sake meter value of −19.6 is most likely due to this high isomaltose concentration. The genomic DNA of MC87-46 encodes for isomaltases IMA1, IMA2, IMA3, IMA4 and IMA5, as well as the isomaltose transporter gene, AGT1. However, these genes were not induced in MC87-46 by isomaltose, and the strain did not possess isomaltase activity. These results show that MC87-46 cannot utilize isomaltose, resulting in its accumulation in the fermented sake. Isomaltose concentrations in sake brewed with MC87-46 were 24.6-fold more than in commercial sake. These findings suggest that MC87-46 may be useful for commercial application in Japanese sake production because of its unique flavour and nutrient profile.

Quantification of organic acids and aroma components in fermented sake. The concentrations of the major organic acids in the sakes were quantified using HPLC (Table 1). Malic acid, succinic acid, lactic acid, and acetic acid were detected in sake produced by both strains. Among them, only acetic acid was increased 1.9-fold in sake fermented with MC87-46 than in that fermented with K901 (Table 1). This suggests that acetic acid was involved in the acidity level of sake brewed with MC87-46.
The aroma components in the sakes were also quantified using headspace gas chromatography. Isoamyl alcohol and ethyl caproate produced by MC87-46 were similar to those produced by K901 (Table 1). The isoamyl acetate and ethyl acetate levels in the sake produced by K901 were 3.70-and 2.85-fold higher, respectively, than those in the sake  produced by MC87-46 (Table 1). Compared with the wild S. cerevisiae MC strain isolated from the carnation flowers, MC87-46 produced 7.0-fold more ethyl caproate (Fig. S1). In addition, a mutation in FAS2 gene at 3748 position was identified by DNA sequencing (data not shown). These results indicated that the ethyl caproate production ability of MC87-46 was improved by selecting for the cerulenin-resistant colony after EMS mutagenesis.
Quantification of sugars in fermented sake. Sugars in sake fermented by both MC87-46 and K901 strains were identified by thin-layer chromatography (TLC) and gas chromatography-mass spectroscopy (GC-MS) (Fig. S3), and quantified using a reducing-sugar high-performance liquid chromatography (HPLC) analytical system (Fig. 2). Sugars in the sake fermented by MC87-46 were separated using TLC (data not shown), and a spot with the same retention time of the isomaltose standard was identified. The mass spectrum of trimethylsilyl (TMS) derivatives from the spot indicated characteristic features of isomaltose fragmentation (Fig. S3). Glucose and panose were also identified by the same procedure (data not shown). Glucose, isomaltose, and panose were detected in sake produced by both strains (Fig. 2). Although the amount of glucose was similar in both batches of sake, levels of isomaltose and panose were 24.6-and 12.8-fold higher, respectively, in sake fermented by MC87-46 than that by K901 (Table 2). Interestingly, the concentration of isomaltose in the sake brewed with MC87-46 was 44.3 mM, indicating that the low sake meter value (−19.6) of the sake brewed with MC87-46 was caused by the high concentration of sugars, including isomaltose (Table 2), and the low concentration of ethanol (Table 1). The relative sweetness value of isomaltose is 54 compared with the 100 of glucose 27 . This suggests that the sake brewed with MC87-46 is sweet owing to the high content of isomaltose (44.3 mM). Furthermore, the time course of the concentrations of isomaltose, panose, and glucose was investigated during the fermentation process (Fig. 3A,B). The decrease in amount of glucose was similar for both MC87-46 and K901 (Fig. 3C). The initial concentration of isomaltose was 44.3 mM and was unchanged until 8 days into the fermentation of the sake by both strains (Fig. 3B). After brewing for 18 days, the isomaltose concentration in sake fermented by K901 was only 4% of the levels found in MC87-46-brewed sake, indicating that MC87-46 could not utilize isomaltose.
There is the possibility that yeast strain with poor fermentation ability accumulates high concentration of sugars, including isomaltose. To clarify this possibility, the time course of sugar metabolism using X2180 strain, which shows low fermentation activity and high isomaltase activity, was determined (Fig. S4). The decrease in amount of glucose and isomaltose by X2180 was similar to that observed with K901 (Fig. S4). These indicate that poor ethanol fermentation ability of MC87-46 does not cause the isomaltose accumulation during the fermentation process.
Sensory evaluation and taste sensor analysis of brewed sake. We performed the sensory analysis of sake brewed with S. cerevisiae MC87-46 and K901 strains. Taste sensor analysis of the five gustatory attributes revealed that sourness of sake brewed with S. cerevisiae MC87-46 was significantly higher than that with K901 (Table 3). These correlated well with the higher acidity in the sake produced by MC87-46 than that in the sake produced by K901. Sensory evaluation showed clear differences in sweetness and sourness between sake brewed with MC87-46 and K901 (Table 3). These indicated that the isomaltose assimilation ability of S. cerevisiae affects the sweetness of the brewed sake. There was no significant difference in other gustatory attributes.  Figure 2. HPLC analysis of sugars in sake fermented by MC87-46 and K901. Sugars in sake fermented by either MC87-46 or K901 at 15 °C for 18 days were quantified using a reducing-sugar HPLC analytical system. Peak 1, panose; peak 2, unknown; peak 3, isomaltose; peak 4, glucose.  (Fig. 4). The growth of MC87-46 and K901 strains in YN media containing glucose or sucrose as the sole carbon source did not differ significantly (Fig. 4). Compared with the MC87-46 strain, maltose was the preferred sugar for growth of the K901 strain (Fig. 4). The MC87-46 strain barely grew in the YN medium 28 containing 1.0% isomaltose as the sole carbon source (Fig. 4).
Heterologous expression of Ima1, the major isomaltase in S. cerevisiae in Escherichia coli expression system was previously reported 28 . Recombinant Ima1 enzymes of K901 and MC87-46 were expressed in E. coli and purified (Fig. S7A). To evaluate the enzymatic activity of the recombinant Ima1 enzymes, respectively, the catalytic conversions of isomaltose were analysed (Fig. S7B). The substrate was efficiently converted by each Ima1, with generation of glucose (Fig. S7B). The specific activity of Ima1 of MC87-46 was similar to that of K901 (Fig. S7C), indicating that the difference between the strains in their ability to assimilate isomaltose was not due to the enzymatic property of Ima1.

MC87-46 K901
Taste sensor analysis  Table 3. Taste sensor analysis and sensory evaluation of sake brewed with K901 and MC87-46. Taste intensities were evaluated on a scale of 1 (weak) to 5 (strong) for the body (1 (thin) to 5 (thick)) and sweetness (1 (dry) to 5 (sweet)). Data are presented as mean values ± standard deviation (error bars) of three independent experiments. *p < 0.01, Student's t test.
Cell extracts from MC87-46 and K901 incubated for 2 and 4 days with isomaltose were prepared to determine if the yeasts possessed isomaltase activity (Fig. S9). MC87-46 and K901 were pre-incubated in YM medium containing 1.0% glucose, and then transferred to YN medium with 1.0% isomaltose. Although pyruvate dehydrogenase (PDH) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) activities were similar in both strains, isomaltase activity of MC87-46 was extremely low compared with that in K901 (Fig. S9). Initial rates of isomaltose hydrolysis by isomaltase were calculated using HPLC as decreases in the isomaltose concentration. The specific activity of isomaltase in cell extracts prepared from K901 was 78.4 U/mg protein, while MC87-46 did not show any enzyme activity. Even though MC87-46 possesses five IMA genes, the strain could not produce active isomaltases. Isomaltase activity in MC87-46 cells correlated well with the non-induction of IMA1 gene by isomaltose.

Quantification of isomaltose in commercially available sake. Isomaltose concentrations in commer-
cially available low-alcohol sake were determined (Table 4). Sake brewed with MC87-46 contained the highest isomaltose and panose levels compared with the commercial sake (Tables 2, 4). The isomaltose assimilation ability of the yeast is attractive for production of sake with new tastes.

Discussion
In the current study, we isolated wild S. cerevisiae MC strain from the carnation flower and produced sake containing a high concentration of isomaltose using a cerulenin-resistant mutant MC87-46 strain. The sake meter value was −19.6 and sensory evaluation showed clear difference in sweetness of brewed sake, indicating that the sweetness of the sake fermented by MC87-46 was comparable to that of sweet dessert wines.
In Japan, the filamentous fungi A. oryzae has been used for over 1,000 years in the traditional food industry 29 . A. oryzae are generally accepted as being safe in food by the USA Food and Drug Administration (FDA) and they have been applied to produce fermented foods, including Japanese rice wine. Starch in rice is hydrolysed to glucose by A. oryzae, and the glucose is converted to ethanol by S. cerevisiae during sake brewing 1 . In addition to glucose, A. oryzae converts starch to various sugars including oligosaccharides such as maltose, isomaltose, panose, and others, which are also found in the sake 4,15-17 . These oligosaccharides are produced by α-amylases and α-glucosidases during sake mash making [15][16][17] . Alpha-glucosidases (EC 3.2.1.20) catalyse liberation of   glucose from non-reducing ends of α-glucosides, α-linked oligosaccharides, and α-glucans 30 . Theoretically, α-glucosidase, belonging to the glycoside hydrolase family 31, is capable of catalysing transglycosylation by a retaining reaction mechanism 15,30 . A. nidulans α-glucosidase AgdB catalyses formation of α-1,6-glucosidic linkages in addition to hydrolysis of starch, resulting in production of isomaltose and panose from maltose 15,16 . In our study, the initial concentration of isomaltose in the sake was about 44.3 mM, indicating that isomaltose was generated by A. oryzae during manufacturing of rice koji. Isomaltose concentration in sake fermented by MC87-46 was 24.6-fold more than that in the sake fermented by K901. The S. cerevisiae BY4741 strain can assimilate isomaltose by α-1,6-glucosidase Ima 28,29 . Five IMA genes (IMA1-5) of S. cerevisiae are located on different chromosomes with 65% to 99% high sequence identity (YGR287c, YIL172c, YJL216c, YJL221c, and YOL157c). Ima shows high affinity for disaccharide isomaltose and palatinose with α-1,6-linkage, and IMA1 encodes the major isomaltase in S. cerevisiae 28 . In addition, growth of S. cerevisiae on isomaltose requires the presence of AGT1, which encodes an α-glucoside transporter 28 . These genes were not induced by isomaltose in MC87-46, and the strain did not possess isomaltase activity. These results indicated that MC87-46 could not utilize isomaltose, resulting in a high concentration of isomaltose in sake fermented with MC87-46.
Isomalto-oligosaccharides (IMOs), including isomaltose, are a heterogeneous group of glycosyl saccharides with an α-1,6-linkage. These oligosaccharides are used in the food, pharmaceuticals, and cosmetics industries due to their unique prebiotic properties [30][31][32][33][34] . IMOs competitively inhibited rat small intestinal α-glucosidases, such as sucrase, maltase, and glucoamylase, and reduced the rate of hydrolysis of sucrose and other α-glucosyl saccharides, including maltose, dextrin, or soluble starch 31 . These oligosaccharides selectively promote the growth of Bifidobacterium in a dose-dependent manner in the human intestine [31][32][33] . Beside their prebiotic effect, IMOs have beneficial effects on blood cholesterol regulation, immunomodulation, and prevention of various diseases 34 . The MC87-46 strain would be useful for preparing traditional Japanese sake with additional nutrient functions.
In summary, we used a S. cerevisiae MC87-46 isolated from the carnation flower, with subsequent introduction of cerulenin-resistance, to produce sake. The brewed sake was sweet due to high levels of the oligosaccharide isomaltose, which was 24.6-fold more than that found in the commercial sake that we tested. These findings suggested that MC87-46 could be useful for commercial application in Japanese sake production, providing a new taste and nutrient functions.
Isolation and identification of wild S. cerevisiae. S. cerevisiae was isolated from the carnation flower (Dianthus caryophyllus L) grown in an experimental farm at Meijo University with the use of selective medium, as previously described 35 . Flowers were preserved in liquid Yeast Nitrogen Base medium (6.7 g/L) with raffinose (6.6 g/L) and ethanol (80 mL/L), then incubated at 28 °C for 72 h. Raffinose was used for the isolation of S. cerevisiae because Saccharomyces species are capable of utilizing raffinose for growth. Aliquots of 100 μL from the culture were plated onto YPG agar (10 g/L yeast extract, 10 g/L peptone, 10 g/L glucose, 20 g/L agar) supplemented with chloramphenicol (30 mg/L) to inhibit bacterial growth. Yeast colonies were isolated after incubation at 26 °C for 3-5 days.
Sequencing of the ITS1/ITS2 region of the 18S rRNA gene was used to identify the strain of S. cerevisiae. DNA was extracted from the yeast cells as described 23,36 and DNA fragments encoding the 18S rRNA gene were amplified with gene-specific primer sets (Table S1). The wild S. cerevisiae MC strain isolated from the carnation flower was used to brew sake in this study.

Selection of cerulenin-resistant S. cerevisiae strain.
A cerulenin-resistant S. cerevisiae strain was generated using a procedure previously described 24 . Briefly, the wild S. cerevisiae MC strain was cultured in 5 mL of YPG medium at 30 °C for 24 h. The culture was centrifuged at 3,000 × g for 5 min, and the cell pellet was washed with sterile water. The yeast cells were suspended in 5 mL of 0.2 M phosphate buffer (pH 8.0), 0.25 mL of 40% glucose, and 0.2 mL of ethyl methanesulfonate (EMS), and incubated at 30 °C for 1 h. The mutagenized yeast cells were washed with sterile water, plated on YPG agar medium containing 50 μM cerulenin and incubated at 30 °C for 7 days. The strain that had the largest colony diameter was selected as the cerulenin-resistant MC87 strain.

Selection of an ethanol-resistant strain.
The cerulenin-resistant MC87 strain was cultured in YPG medium at 30 °C for 24 h, centrifuged, and resuspended in acetate buffer (pH 4.2). The cell suspension was centrifuged at 3,000 × g for 5 min, and the supernatant was discarded. The cell pellet was again resuspended in acetate buffer (pH 4. Small sake brewing. Small-scale sake brewing was carried out as described previously with some modifications 3 . S. cerevisiae MC87-46 and S. cerevisiae Kyoukai No. 901 (K901) obtained from the Brewing Society of Japan were cultured in YPG medium at 30 °C for 2 days and then harvested by centrifugation at 200 × g. The small-scale sake brewing test included 10 mL of pre-incubated yeast culture (1 × 10 6 cells/mL), 40 g of dried rice (Tokushima Seikiku, Osaka, Japan), 10 g of rice koji (Tokushima Seikiku), and 90 mL of water mixed in 250 mL glass bottles. Fermentation proceeded for 18 days at 15 °C. The weight of the sake mash was measured to monitor carbon dioxide evolution. Prepared sake (100 mL) was separated by centrifugation at 8,000 × g for 5 min, and the resulting supernatant was filtered through a 0.22-μm filter.
Analytical methods. Sake meter value, sake acidity, and amino acid levels were measured as previously described 6 . Organic acids were quantified using HPLC with a Shim-pack SCR-102H column (Shimadzu, Kyoto, Japan). Aroma components were quantified using headspace GC with a J & W DB-WAX capillary column (30 m × 0.32 mm internal diameter × 0.50 μm film thickness; Agilent Technologies, Santa Clara, California, USA). Ethanol concentration was measured using an ethanol analyser (RIKEN KEIKI, Tokyo, Japan). Sugars were determined by monitoring post-column derivative reducing sugars separated using a Prominence reducing-sugar HPLC analytical system (Shimadzu) equipped with a fluorescence detector. Sake fermented with MC87-46 and K901 was separated on a Shim-pack 4.0 × 250-mm ISA-07/S2504 column (Shimadzu) with a linear gradient of 0.1 M potassium borate buffer (pH 8.0) and 0.4 M potassium borate buffer (pH 9.0) for 140 min at a flow rate of 0.6 mL/min 37,38 . Sugars were also lyophilized, trimethylsilylated, and analysed using a GCMS-QP2010 (Shimadzu) equipped with a J & W DB-5MS capillary column (30 m × 0.25 mm internal diameter × 0.25 μm film thickness; Agilent Technologies) 39 .
Sensory evaluation and taste sensor analysis. The taste of sake was evaluated by ten students, belonging to Japanese sake study club of Meijo University, experienced in sensory evaluation of sake. The samples were evaluated for six attributes (sweetness, umami, sourness, bitterness, aftertaste, body) and scored on a scale from 1 (weak) to 5 (strong) for body (1 (thin) -5 (thick)) and sweetness (1 (dry) -5 (sweet)) 26 .
A taste sensing apparatus (SA402B, Intelligent Sensor Technology Inc., Kanagawa, Japan) was used for the taste sensor analysis, according to the manufacturer's instructions.

Ethical permission. All experiments were performed according relevant guidelines and regulations of Meijo
University. The Research Commission of the Faculty of Agriculture at Meijo University approved the procedures and methodologies related to the sensory evaluation of Japanese sake brewed with K901 and MC87-46 strains. All participants of the sensory evaluation tests signed an informed consent form on the type of the research being conducted; they agreed to sensory evaluation of the Japanese sake that was produced as described in the Methods section under "Small sake brewing. " Yeast cell growth using various sugars. MC87-46 and K901 (1 × 10 5 cells) were inoculated in YN medium supplemented with glucose, sucrose, maltose, or isomaltose (1.0% each) at 30 °C with shaking at 120 rpm 28 . Cell growth was monitored by optical density at 600 nm.
Cloning of IMA1-5 and AGT1 genes. Total RNA was prepared from disrupted yeast cells using the RNeasy Mini Kit (Qiagen, Venlo, The Netherlands), according to the manufacturer's instructions. Single-stranded cDNA was synthesized from total RNA extracted, and cDNA fragments encoding the full-length IMA1-5 and AGT1 genes were prepared by PCR using gene-specific primer sets (Table S1). The PCR product was cloned, and the nucleotide sequence was determined using an automated DNA sequencer (CEQ2000, Beckman Coulter, Brea, CA, USA).
Preparation of recombinant proteins. The PCR product of IMA1 gene was purified, and then ligated into pET21a (Novagen, Darmstadt, Germany), which had been digested with the restriction enzyme BamHI. Plasmid for producing recombinant Ima1 of K901 was constructed by the same strategy.
The plasmids were introduced into E. coli BL21-CodonPlus(DE3) (Novagen), cultured in LB containing 100 μg/L sodium ampicillin for 12 h, and then a portion (3 mL) was subjected to rotary shaking at 120 rpm in 200 mL of LB containing 100 μg/L sodium ampicillin at 30 °C. Isopropyl-thio-β-d-galactoside (0.2 mM) was added to the medium when the OD 600 reached 1.0, and the flasks were rotary-shaken for 8 h at 120 rpm and 30 °C. The E. coli cells were harvested, suspended in 50 mL of 20 mM sodium phosphate buffer (pH 7.0) containing 1 mM EDTA and 0.1 mM PMSF, and disrupted by sonication. The sonicate was centrifuged at 6,000 × g for 10 min, and the soluble fraction was separated by centrifugation at 20,000 × g for 15 min. These fractions were passed through a column containing nickel-nitrilotriacetic acid agarose (φ5 × 20 mm, Qiagen) that was washed with 10 mL of 20 mM sodium phosphate (pH 7.0) containing 20 mM imidazole. Proteins were eluted with the same buffer containing 300 mM imidazole. All procedures were carried out at temperatures below 4 °C. The resultant fraction was dialyzed against 20 mM sodium phosphate (pH 7.0) and analysed.
Enzyme assays. MC87-46 and K901 cells (OD 600 nm = 1.0) grown in 10 mL of YN medium containing 1.0% glucose (w/v) were collected by centrifugation (2,000 × g, 5 min). The cell pellets were washed twice with 1.0 mL of Milli Q water and transferred to a test tube containing 10 mL of YN medium with 1.0% isomaltose (w/v). After incubation at 25 °C with shaking at 120 rpm, cells were collected and resuspended in 700 μL of 0.1 M potassium