2022 年 28 巻 3 号 p. 217-224
Sake yeasts synthesize alcohols and other components, such as ethyl caproate, ethyl caprylate, and isoamyl acetate, which impart a unique and refined flavor profile to sake. In this study, a sake yeast mutant, K901C8, with high ethyl caprylate productivity was isolated from Kyokai no. 901 (K901), which has been passaged since 1993 at Yoshinogawa brewery, and its fermentative properties were investigated. This mutant was isolated from 960 antibiotic-resistant mutants obtained from 2.1× 107 cells with cerulenin as an inhibitor of fatty acid synthesis. In small-scale sake brewing experiments, the K901C8 mutant produced a high level of ethyl caprylate, which imparts a pineapple- and apricot-like flavor to sake. In the industrial-scale brewing, K901C8 showed fermentative properties similar to those of the wild-type K901, whereas K901C8 produced more ethyl caprylate than K901. These results demonstrate the potential of using K901C8 for producing high-quality sake with a novel flavor.
During fermentation of alcoholic beverages, such as wine, beer, whiskey, and sake, the yeast Saccharomyces cerevisiae synthesizes various aroma and volatile flavor compounds. Among these, fatty acid esters are by far the most prevalent group of flavor compounds in liquors produced via yeast fermentation. Several fatty acid esters, such as medium-chain fatty acids containing ethyl caproate and ethyl caprylate, are important for the development of alcoholic beverages owing to their contribution to beverage aroma (Abbas, 2006).
Sake is a traditional Japanese alcoholic beverage brewed using steamed rice, koji (koji mold, produced by Aspergillus oryzae cultured on steamed rice), and sake yeast (S. cerevisiae) (Yoshizawa, 1999). During sake production, sake yeasts play an important role in synthesizing not only alcohols but also other components, such as ethyl caproate, ethyl caprylate, and isoamyl acetate, that impart the unique and refined flavor profile commonly associated with sake (Ginjo-ka). Numerous strains of yeast have been studied or engineered to improve the flavor of sake (Kitagaki and Katsuhiko, 2013). For example, a cerulenin-resistant yeast strain that produces high levels of ethyl caproate is widely used in sake breweries (Ichikawa et al., 1991). Cerulenin is an antifungal antibiotic that inhibits the biosynthesis of fatty acids and steroids. This mutant carries a point mutation (3748G>A) in FAS2, a gene encoding fatty acid synthase (FAS). This mutation results in an amino acid change (FAS2-G1250S) that confers a high caproate-secreting phenotype, resulting in the production of high levels of ethyl caproate (Inokoshi et al., 1994; Akada et al., 1999). In S. cerevisiae, FAS catalyzes the biosynthesis of fatty acids to form molecules of specific lengths for use in the whole metabolic pathway. This well-characterized, multifunctional enzyme (type I system) consists of six alpha and six beta subunits (α6β6 complex) encoded by FAS2 and FAS1, respectively. These genes control de novo biosynthesis of fatty acids; therefore, the FAS2 locus of sake yeast is an important target for improving the flavor of sake during fermentation.
In recent years, new mutations in the FAS2 gene have been identified in S. cerevisiae, and these mutations confer the yeast with the ability for high caprylic acid production. Gajewski et al. (2017) reported that the FAS2-R1834K mutation causes a remarkable increase in the level of extracellular short-chain fatty acids—mostly caprylic acid. When the introduction of several FAS2 mutations (I306A, G1250S, M1251W, F1297Y, and R1834K) was tested in different combinations, the triple mutant I306A-R1834K-F1297Y was found to specifically enable the enhanced production of caprylic acid. Likewise, genetic engineering technology was used to isolate a mutant strain harboring a point mutation (FAS2-F1279Y) that produces high levels of caprylic acid in S. cerevisiae (Xue et al., 2020). On the contrary, we also used a traditional screening method (Ichikawa et al., 1991) to isolate a mutant yeast strain harbouring a point mutation (FAS2-F1279Y); this mutant produces high levels of caprylic acid in the koji extract medium (Kuribayashi et al., 2019). Similarly, our teams have previously reported that the isolation of a novel cerulenin-resistant mutant that produces high levels of caprylic acid, leading to elevated levels of ethyl caprylate that result in a unique pineapple-like aroma in the koji extract medium (Nagai et al., 2016). This strain harbours a point mutation nucleotide (3758G>C) in FAS2 that results in a change in amino acid residue 1253 (FAS2-G1253A) of the α-subunit. These mutants appear to increase levels of ethyl caprylate, conferring a unique fruity aroma and sweet flavor. However, there is no information about the food and beverage fermentative properties of yeast strains that produce high levels of ethyl caprylate and sake has not yet been produced with such mutants.
In this study, we isolated and assessed the fermentation properties of a novel S. cerevisiae mutant derived from the industrial sake yeast strain Kyokai no. 901 (K901), which had been subcultured for a long time in Yoshinogawa brewery, using the cerulenin screening method. Our isolated mutant, K901C8, produces higher levels of caprylic acid/ethyl caprylate than the parent strain K901. Furthermore, the fermentative properties of this mutant were assessed at an industrial brewing scale, resulting in the demonstration of production for high-quality sake with a unique flavor profile.
Yeast strain One of the most widely used industrial sake yeast strains, K901, was obtained from the Brewery Society of Japan in 1993. It has been successively subcultured for approximately 24 years on koji extract medium (Ishiyama et al., 2008) at Yoshinogawa Co, Ltd. (Niigata, Japan). The yeast strain was grown on an agar-slant medium containing koji extract, stored at 4°C, and transferred every 6 months onto the same slant medium to maintain cell viability.
Isolation of cerulenin-resistant sake yeast mutants To isolate mutants, the sake yeast strain K901 was inoculated onto Yeast Extract-Peptone-Dextrose (YPD) medium (1% yeast extract, 2% polypeptone, and 2% glucose) and incubated at 25 °C for 5 days under static conditions. Cells were harvested, washed five times with sterilized water, and spread at a density of 1.0 × 106 cells/plate on a solid medium containing 0.17% yeast nitrogen base without amino acids (Difco; Detroit, MI, USA), 1 µg/L cerulenin (Sigma Aldrich, St. Louis, MO, USA), 2% glucose, and 2% agar. The plates were incubated at 30 °C for 7 days. Yeast colonies growing on this selective media were cerulenin-resistant mutants (Ichikawa et al., 1991). Screening to detect mutants with high productivity of free fatty acids (FFAs) was carried out by high-throughput screening, as previously described (Kuribayashi et al., 2012). Briefly, the isolated cerulenin-resistant mutants were inoculated onto YPD medium in a 96-well microplate and incubated at 30 °C for 2 days. The FFA concentration in each well was examined using the NEFA C-test kit (Wako Pure Chemical Industries Ltd., Osaka, Japan), and one mutant showing high FFA productivity was selected; this mutant was designated K901C8.
Free fatty acid analysis To analyze the FFAs produced by the mutant, yeast strains were cultured in a koji extract medium. Yeast cells were inoculated into 300 mL of koji extract medium (2.5×105 cells/mL) and cultured at 25 °C for 7 days without shaking. The culture supernatant was filtered through 0.45 µm cellulose acetate membrane cartridge filters (ADVANTEC, Tokyo, Japan), and the filtrate was analyzed by gas chromatography (GC), as described by de Jong and Bading (1990) and Kuribayashi et al. (2012). Briefly, 2 g of sodium chloride and 0.15 mL of sulfuric acid were added to 50 ml of the filtrate. The FFAs were extracted by adding 50 mL of the mixture of ether/heptane (1:1, v/v) containing 0.5 mg of heptanoic acid as an internal standard and then cleaned using an NH2 cartridge (500 mg; GL Sciences Inc, Tokyo, Japan). The following conditions were used for GC analysis: instrument, GC-14B (Shimadzu, Kyoto, Japan); detection, flame ionization; column, InertCap FFAP fused-silica capillary type (0.25 mm I.D. × 15 m, df = 0.25 mm, GL Sciences); column temperature, 100 °C for 5 min, rising to 240 °C at 10 °C/min, and then held for 20 min; carrier gas, helium at 50 kPa.
Sake brewing analyses To examine the fermentation properties of K901C8, a sake brewing test was carried out using K901 as the control strain. First, the production of ethyl caprylate by K901C8 was investigated by performing a small-scale brewing as described by Namba et al. (1978) using 160 g of pre-gelatinized rice (Tokushima-seikiku, Tokushima, Japan) with a polishing ratio of 60%, 40 g of dried koji (Tokushima-seikiku) with a polishing ratio of 70%, and 300 mL of water with 0.03% (v/v) lactic acid per batch. The fermentation mash was maintained at 15 °C. Fermentation profiles were monitored by measuring CO2 release and fermentation was terminated when the total released CO2 reached approximately 60 g/mash. After fermentation, the sake mash was centrifuged, and the supernatant was obtained as sake.
An industrial-scale sake brewing test was carried out using a total of 100 kg of rice (steamed rice and koji); 80 kg of rice with a polishing ratio of 65% was used for the steamed rice and 20 kg of rice with a polishing ratio of 65% was used to prepare koji. On the first day, yeast (6× 1011 cells) was suspended in 30 L of water; 5 kg of koji and 90 mL of 90% (v/v) lactic acid were added to the yeast cell suspension at 15 °C (Table 1). The next day, 15 kg of steamed rice was added to the mash at 15 °C. For yeast growth, the mash was incubated without thermal management for one day. On the third day, 25 kg of steamed rice, 5 kg of koji, and 50 L of water were added, and the fermentation was carried out at 10 °C. On the fourth day, 40 kg of steamed rice, 10 kg of koji, and 60 L of water were added, and fermentation was continued at 9 °C. Subsequently, the sake mash was fermented for 6 days with the temperature increased by 1.0 °C per day to a maximum temperature of 15 °C. The sake mash was periodically sampled during fermentation and subjected to component analysis. After fermentation, the sake mash was centrifuged, and the supernatant was obtained as sake. The general properties, aromatic and flavor compounds, and the concentrations of caproic acid and caprylic acid in the sake produced after each test were analyzed using head-space gas chromatography (GC-14A; Shimadzu, Kyoto, Japan) following a standard method established by the National Tax Agency of Japan (Okazaki, 1993).
| Materials | Seed mash (Shubo) | Sake mash (Moromi) | Total | ||
|---|---|---|---|---|---|
| 1st (Soe) | 2nd (Naka) | 3rd (Tome) | |||
| Rice for steaming (kg) | 0 | 15 | 25 | 40 | 80 |
| Rice for koji (kg) | 5 | 0 | 5 | 10 | 20 |
| Water (L) | 30 | 0 | 50 | 60 | 140 |
| 90%(v/v) Lactic acid (mL) | 90 | 0 | 0 | 0 | 90 |
Sensory evaluation Sakes brewed at an industrial-scale using K901C8 and K901 were subjected to sensory evaluation according to the method described by Takahashi et al. (2014). Briefly, sensory evaluations were performed by ten trained sensory panelists from Yoshinogawa brewery using the flavor profile method (Kobayashi, 1989). The panelists were informed about the characteristics of these yeast strains, and the ethyl caprylate-associated aroma characteristics were evaluated; these were identified as “pineapple-like.” After tasting the sake, the panelists were asked to describe its characteristics. In this way, the ethyl caprylate aroma characteristics were evaluated and identified in sake.
This study was approved by the Research Ethics Committee of the Niigata Agro-Food University (Approval No: 8-211203; date of approval 03-12-2021).
Statistical analysis Data obtained in the small-scale brewing test were expressed as mean ± SD and compared using Student's t-test. As a note, since the industrial-scale sake production was carried out only once, the data was not statistically analyzed.
Isolation and characterization of S. cerevisiae K901C8 Treatment with mutagens, such as ethyl methanesulfonate and ultraviolet light, has been widely used to generate industrial yeast strains. Although these mutagens efficiently induce targeted mutations, there is also a risk of other genes being mutated, changing the fermentative properties of the parent yeast. To avoid this, we tried to isolate spontaneous mutants arising from industrial sake yeast K901 which has been passaged since 1993 at Yoshinogawa Brewery.
In the first screen for mutants using the cerulenin plate, 960 cerulenin-resistant mutants were obtained from 2.1× 107 cells. A second high-throughput screen yielded only two spontaneous mutants with high FFA productivity. One mutant (K901C6) harbored the FAS2-G1250S mutation, which has been reported to confer the ability to produce ethyl caproate (Inokoshi et al., 1994; Akada et al., 1999). As shown in Fig. 1, the levels of caproic acid and caprylic acid secreted by K901C6 were higher than those secreted by the parental strain K901 in the koji extract medium (K901C6; caproic acid: 7.4 mg/L and caprylic acid: 3.7 mg/L). Another mutant, K901C8, produced more than 13-fold higher levels of caprylic acid than the K901 strain (K901C8: 5.3 mg/L vs. K901: 0.4 mg/L, respectively). On the other hand, the concentration of caproic acid produced did not show a notable change, compared with that produced by K901C6 (K901C8: 1.7 mg/L vs. K901: 0.7 mg/L, respectively). This finding shows that K901C8 has a new mutation, which is different from that of K901C6. Indeed, sequence analysis of the FAS2 locus of K901C8 confirmed that this mutant harbored a homozygous FAS2 3836T>A point mutation, which altered the amino acid sequence of FAS2 (Fas2-F1279Y) (Fig. 1).
The FAS2 gene mutation in the Saccharomyces cerevisiae K901C8 mutant and analysis of free fatty acids in koji extract medium via gas chromatography. The black triangle points to the location of the mutation in the FAS2 gene (T3836A), which alters the phenylalanine at residue 1279 to a tyrosine (Fas2-F1279Y) of the FAS2 enzyme. Meanwhile, the white triangle represents to the position of the mutation in the FAS2 gene (G3748A), which changes the glycine at residue 1250 to a serine (Fas2-G1250S) of this protein. In the panels of gas chromatography, “I.S.” indicates the internal standard (10 mg/L of enanthic acid). 6:0, caproic acid; 8:0, caprylic acid.
In this isolation experiment, two mutants were obtained from sake yeast strain K901, and one of them harbored a new mutation FAS2-F1279Y. One reason for the emergence of the unique mutant could be the long-term passaging of the parental strain. Takashita et al. (2013) also reported that useful mutants for shochu (Japanese distilled liquor) production were isolated from a brewer's yeast strain by subculturing for a long period of time in barley shochu mash or in conventional complete medium such as YPD broth. Although more detailed investigation is required, these findings suggest that subculturing for a long time may be useful as a tool for obtaining novel yeast strains.
In S. cerevisiae, FAS catalyzes the biosynthesis of fatty acids to form molecules of specific lengths that can be used in the lipid metabolic pathway. The α-subunit consists of multiple domains, including the acyl carrier protein, ketoacyl reductase, ketoacyl synthase (KS), and phosphopantetheine transferase. The KS domain performs the key condensation reaction to elongate the fatty acid chain and is targeted by cerulenin (Johansson et al., 2008). Xue et al. (2020) have previously shown that FAS2-F1279Y mutation induced by genetic transformation led to enhance the production of caprylic acid in S. cerevisiae. Hence, the enhanced caprylic acid in strain K901C8 can be explained by the FAS2-F1279Y mutation. However, further study is required to understand the role of this mutation in the activity of the S. cerevisiae fatty acid synthase.
Sake brewing with K901C8 A small-scale brewing test and industrial fermentation were carried out to investigate the fermentative properties of K901C8.
The characteristics of the sake obtained by brewing with K901C8 and K901 are shown in Table 2. Notably, the general properties (alcohol, acidity, and amino acidity) of sake produced by K901C8 did not significantly differ from sake produced by K901. However, the sake meter of K901C8 was significantly lower than that of K901 because of a slight delay in fermentation. The concentration of caprylic acid in the sake was significantly higher in K901C8 than in K901 (3.7-fold, p <0.05), whereas the concentration of caproic acid did not differ between sakes produced by the two strains. Likewise, K901C8 produced approximately 4.0-fold higher amounts of ethyl caprylate in the sake than did K901 (Table 2).
| Yeast strain | General properties | Free fatty acids (mg/L) | Flavor components (mg/L) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| CO2 evolution (g) | Sake meter | Alcohol (%) | Acidity (mL) | Amino acidity (mL) | Caproic acid | Caprylic acid | Ethyl caproate | Ethyl caprylate | |
| K901 | 59.9 ± 1.1 | −33 ± 1 | 17.7 ± 0.3 | 3.8 ± 0.0 | 1.5 ± 0.1 | 2.7 ± 0.0 | 1.5 ± 0.1 | 0.9 ± 0.0 | 0.4 ± 0.0 |
| K901C8 | 59.7 ± 0.6 | −42 ± 0* | 17.1 ± 0.0 | 3.8 ± 0.1 | 1.8 ± 0.1 | 2.9 ± 0.2 | 5.6* ± 0.2 | 1.0 ± 0.1 | 1.6 ± 0.1 |
Furthermore, in an industrial-scale (100 kg rice) brewing test, K901C8 showed a similar fermentation profile to K901 (Fig. 2). The fermentation times of sake mash (moromi period) for K901C8 and K901 were the same (22 days). Moreover, the general properties of K901C8 did not differ from those of K901 during sake production and the final products were similar (Table 3). The amounts of caprylic acid and ethyl caprylate produced by K901C8 were 4.0- and 3.4-fold higher than those produced by K901, respectively. Interestingly, K901C8 produced lower amounts of other flavor components, including isobutyl alcohol, isoamyl alcohol, and isoamyl acetate, than K901.
Industrial-scale brewing test using K901 and K901C8. (a) The amount of alcohol and (b) values of sake meter in the sake mash.
| Yeast strain | General properties | Free fatty acids (mg/L) | Flavor components | (mg/L) | Sensory scorea | Number of panelists detecting a pineapple-like flavorb | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Sake meter | Alcohol (%) | Acidity (mL) | Amino acidity (mL) | Caproic acid | Caprylic acid | Isobutyl alcohol | Isoamyl alcohol | Isoamyl acetate | Ethyl caproate | Ethyl caprylate | |||
| K901 | +4 | 17.2 | 2.1 | 1.2 | 3.5 | 3.4 | 55.9 | 2.9 | 148.5 | 1.2 | 0.9 | 3.0 | 0 |
| K901C8 | ±0 | 16.8 | 2.2 | 1.1 | 3.4 | 13.6 | 39.0 | 1.7 | 118.9 | 1.5 | 3.1 | 2.7 | 10 |
In a sensory evolution test, the sake brewed using K901C8 had an improved sensory score than that brewed using K901 (Table 3). This finding indicates that the K901C8 strain has the desirable characteristics for sake brewing at a industrial-scale. In sakes, caprylic acid is recognized as an off-flavor. Yamane et al. (1997) reported that the threshold for caprylic acid in 15% ethanol is 2.4 mg/L. However, the sensory evaluation results of K901C8 sake, which is rich in medium-chain fatty acids, were superior to those of its parent strain; additionally, no bad odor was indicated. This is because the flavor of sake is formed by the balance of various aromas; thus, the flavor balance of K901C8 might mask the flavor of caprylic acid. Since K901C8 produced sake with lower levels of alcohols and isoamyl acetate than K901, it is possible that the flavor composition of K901C8 changed remarkably from that of its parent K901 strain. However, this requires further detailed examination, such as metabolome analysis.
Next, to confirm whether the pineapple-like flavor derived from ethyl caprylate was recognizable in sake produced by K901C8, a sensory evaluation of the brewed sake was carried out with a skilled panelist. In the sensory evaluation, the panel found that the sake brewed by K901C8 had a pineapple-like flavor. In industrial sake brewing, particularly Ginjo-shu (high-grade sake) brewing, cerulenin-resistant mutants that harbor the FAS2-G1250S mutation are widely used. These mutants produce 10.6–42.5 mg/L of caproic acid and 3.4–13.2 mg/L of ethyl caproate (Kuribayashi et al., 2014); thus, the apple-like flavor derived from ethyl caproate is the predominant in the flavor these sakes. By contrast, K901C8 yielded a sake with more caprylic acid than caproic acid and the pineapple-like flavor derived from ethyl caprylate became the main component of its characteristic flavor. This indicates that the K901C8 mutant has potential value for commercial production of sake with a novel flavor.
Moreover, this study demonstrates that the FAS2-F1279Y mutation can be applied to various industrial yeasts to craft alcoholic beverages with unique quality attributes. As an exception, although a new application of Saccharomyces and non-Saccharomyces yeasts for wine brewing demonstrated that improved medium-chain fatty acid ethyl esters specifically enhanced the temperate fruity aroma of wine (Hu et al., 2018), the technique had a low reproducibility. As the brewing industry has become interested in producing a greater diversity of novel alcoholic beverages in recent years, brewers are exploring novel yeast strains for specialty liquor and bio-flavoring (Nikulin et al., 2020) and the results of this study indicate that the FAS2-F1279Y yeast mutant can efficiently promote the development of various alcoholic beverages across the world.
In conclusion, we isolated a new sake yeast strain showing high productivity of ethyl caprylate, resulting in a unique flavor profile of the sake produced using it. Our results will be valuable for further improving sake yeasts for industrial-scale applications. In addition, the fermentation properties of the yeast mutant would provide important information, not only for the production of sake, but also for the manufacture of all alcoholic beverages, such as beer, cider, and wine.
Acknowledgements We thank the Brewing Society of Japan for providing the yeast strains. We also thank all members of Yoshinogawa Co., Ltd. for their assistance during the study. The FAS2-1279Y mutation in S. cerevisiae is the patent of Niigata prefecture in Japan (JP6587087B2). This work was supported by JSPS KAKENHI (Grant Number JP20K22570).
Conflict of interest There are no conflicts of interest to declare.
