Effect of sake lees on cheese components in cheese ripened by Aspergillus oryzae and lactic acid bacteria

More than 2,000 varieties of cheese currently exist in the world, and cheese manufacture continues to flour-ish. To develop the cheese ripening process, additional ingredients are used during cheese production. In this study, the effect of sake lees as an additional ingredient on the fermentation of cheese using Aspergillus oryzae (koji mold), known as koji cheese, was investigated. Aspergillus oryzae is used in the fermentation of Japanese traditional foods, such as sake and soy sauce, given its strong enzymatic activities, as well as in cheese production (i.e., koji cheese). Sake lees, a by-product of the fermentation of rice with A . oryzae and yeasts in the sake brewing process, contains various metabolites, such as amino acids. Here, supplementation with sake lees enhanced the activities of lactic acid bacteria and affected the color of the cheese. Metabolome analysis revealed that sake lees altered the balance of carbohydrates and fatty acids in the cheese. Remarkably, supplementation with sake lees enhanced the production of umami-enhancing γ-glutamyl (kokumi-active) peptides. This study suggests that a new type of cheese can be produced using A . oryzae and sake lees, and information on the synergistic effects of A . oryzae and sake lees will aid the development of cheese production.


INTRODUCTION
The beginning of cheese manufacturing is believed to date back to at least the sixth millennium BC in northern Europe, and techniques for its production have been evolving ever since (Salque et al., 2013).There are various types of cheese, including hard, semi-hard, fresh, and others (Johnson, 2017).The main cheese starter microorganisms, lactic acid bacteria (LAB), have been extensively studied for their glycolytic, proteinase, and peptidase activities, as these lead to the production of organic acids, peptides, and AA that can strongly affect cheese ripening (Fox et al., 1996;McMahon et al., 2014).In addition to LAB, other specific bacteria, and certain fungi such as Brevibacterium linens, Geotrichum candidum, Penicillium camemberti, and P. roqueforti, are also used in cheese making, for ripening, and to generate flavor and texture (Arfi et al., 2003;Kalai et al., 2017).
The fungi used for cheese making generally exhibit high proteinase and lipolytic activities, although activities vary between strains (Larsen et al., 1998;Larsen and Jensen, 1999;Lessard et al., 2014;Mane and Mc-Sweeney, 2020).The concentrations of AA in cheeses can be influenced by the interaction between bacteria and fungi (Hansen and Jakobsen, 1997).Free fatty acids are generated in abundance from the lipids (triacylglycerides) in milk through the lipase activity of fungi.Cheeses made with fungi, especially blue cheeses made using P. roqueforti, contain greater concentrations of free fatty acid, such as octadecanoic acid (C18:0), when compared with cheeses cultured without fungi (Collins et al., 2003).
Aspergillus oryzae and A. sojae, known as koji molds, are used in the fermentation of Japanese traditional products, including sake (rice wine) and shoyu (soy sauce).They exhibit not only proteolytic and lipase activity but also glycosidase activity (Toida et al., 2000;Kum et al., 2015;Ichishima, 2016).A cheese ripened by A. oryzae chosen B was launched in 1962, and its taste and volatile compounds have been investigated (Tokita and Nakanishi, 1962;Nakanishi and Nakazawa, 1965).When ripened by A. oryzae, the taste and variety of volatile sulfur compounds in the cheese differed from those in commercial camembert and blue cheese.Although koji cheese was commercially produced in the 1960s, production ended because of low market demand (Japanese cheese consumption was ~30 million kg/yr).In addition, the Japanese were not familiar with natural cheese.However, in 2020, cheese consumption was >10-fold higher than that in the 1960s, at ~350 million kg/yr (data provided by the Ministry of Agriculture, Forestry and Fisheries of Japan).Modified koji cheese is considered to have become attractive to overseas consumers as well as Japanese consumers, because Japanese traditional food, such as soy sauce and sake fermented with koji, are now available worldwide.
A recent study showed that cheese produced with some koji molds (A.oryzae and A. sojae) had a milder flavor than camembert ripened by P. candidum, because of their low lipase activity and low production of volatile short-chain fatty acids (Suzuki et al., 2021).These studies indicate the fungus A. oryzae could be used for developing novel types of cheese.
The enzymatic activity of microorganisms is one of the most important factors affecting cheese ripening, with flavor compounds being generated through the action of proteinases and lipases.The proteolytic activity of LAB is also needed to generate EAA from casein in the milk (Savijoki et al., 2006).
Various ingredients, such as enzymes from plants or microorganisms, and cheese components, have been added to cheeses to accelerate cheese ripening and develop flavor (Fernandez-Garcia et al., 1994;Hannon et al., 2006;Spelbrink et al., 2015).Additionally, it has been suggested that food-derived ingredients, such as basil essential oil, can promote the growth of starter LAB (Abbas et al., 2018).Sake lees (sake-kasu) is a by-product from the production of sake, a Japanese alcoholic beverage made from rice fermented by koji (A.oryzae) and sake yeast (Saccharomyces cerevisiae).Sake lees is considered useful as a booster for cheese starter cultures and cheese ripening because it contains high concentrations of AA and proteinase, essential for the growth of the starter and the degradation of the milk protein (Tsutsui et al., 1998).In fermented sausage, supplementation with sake lees beneficially decreases pH by promoting the action of LAB (Mikami et al., 2020).These results imply that sake lees could boost the activity of LAB in cheese.Although the effect of sake lees on the A. oryzae and LAB together in cheese is unknown, we surmised that sake lees may enhance cheese development in collaboration with A. oryzae.
In this study, metabolome analysis was conducted to examine the effect of sake lees on koji cheese ripened by A. oryzae, focusing on AA, fatty acids (FA), and peptides.Our results provide new information on cheese ripening by A. oryzae in conjunction with sake lees.

Cheesemaking
The koji cheeses were manufactured from nonstandardized pasteurized milk (fat 3.6%) produced at Zao Dairy Center, Miyagi, Japan.Sake lees (sake kasu, a gift from Hakkaisan Brewery Co. Ltd., Niigata, Japan) was derived from the filtered residue of Japanese sake mash (Kurahashi, 2021).Plain milk or milk with 2% (wt/vol) sake lees was pasteurized at 72°C and then cooled in the manufacturing tank to 30°C.A mesophilic lactic acid culture (freeze-dried Redi-set culture Chr. Hansen Inc.) was used for prematuration (1 U/L milk).Calcium chloride (Japanese food additives, Kanto Chemical Co.Inc.) was added to all recipes at a final concentration of 0.01% (wt/vol) to enhance curd formation.Microbial rennet (MRS, Meito Sangyo Co. Ltd.) was stirred gently into the fermented milk at a rate of 0.06 mL/L of milk.The mixture was allowed to set for 30 to 45 min before cutting with a horizontal and transverse lyre.The curd was then mixed slowly for 2 min, curd washing was performed by replacing 75% (vol/vol) of the whey with water at 30°C, and the curd was allowed to stand for 30 min.For cheese molding, the curd was ladled into cheese hoops (8 cm diameter, 3 cm height), which were then turned 4 times per hour.The cheeses were drained overnight at 30°C.Subsequently, the cheeses were unmolded and soaked in a salt brine (20% wt/vol) for 10 min and kept for 2 h in a refrigerator at 10°C for drying.At this time, preripening samples were collected, named C0 (control cheese without sake lees) and S0 (cheese supplemented with sake lees).For the treatments involving inoculation with the fungus, the cheeses were soaked in koji mold (A.oryzae KC43, Higuchi Matsunosuke Shoten Co. Ltd.) dispersed in sterilized distilled water to a concentration of 2 × 10 6 /mL.The cheeses were then ripened at 35°C for 4 d and then at 30°C for 3 d in a plastic box held at approximately 90% relative humidity to encourage uniform surface mold development.The koji cheeses were then wrapped in plastic film, stored at 10°C, and analyzed 30 d after manufacture.The samples were named C30 (C0 ripened for 30 d) and S30 (S0 ripened for 30 d).The cheeses were made in triplicate, and all 12 samples (3 each for C0, S0, C30, No animals were used in this study, and ethical approval for the use of animals was thus deemed unnecessary.

Bacterial Counts and pH
Before analysis, the rinds were removed from the cheeses.For LAB counts, 5 g of cheese was added to 50 mL 2% sodium citrate and homogenized using IKA T25 digital Ultra-Trrax (12,000 rpm, 1 min at room temperature; IKA Werke GmbH and Co. KG).Each homogenized sample was spread onto an M17 (Difco Laboratories) agar plate supplemented with 0.5% glucose.Plates were incubated at 30°C in an AnaeroPack system (AnaeroPack, Mitsubishi Gas Chemical).For pH analysis, 5 g of cheese was added to 10 mL sterilized distilled water and homogenized.After centrifugation at 12,000 rpm for 10 min, the pH of the supernatant was analyzed using a pH meter (Seven Easy pH, Mettler Toledo).Differences between cheese supplemented with sake lees (n = 3) and control cheese (n = 3) were assessed using t-test.A P-value <0.05 was considered to indicate statistical significance.

Color
Ten grams of cheese was added to 50 mL of distilled water.After homogenization using a Pure Natura mixer (Tescom Denki Co. Ltd.), the color of the suspension was measured according to the CIELAB color system for lightness (L*), redness (a*), and yellowness (b*) using a colorimeter .Differences between the cheese supplemented with sake lees (n = 3) and the control cheese (n = 3) were assessed using Tukey's honestly significant difference test by R 4.0.3(https: / / www .r-project .org/).

Metabolome Analysis
Metabolome analysis was conducted at Human Metabolome Technologies (HMT; Tsuruoka, Japan).Capillary electrophoresis time-of-flight mass spectrometry (CE-TOFMS) was used, as previously described (Hagi et al., 2016).Briefly, a 30-to 35-mg cheese sample was dissolved in 600 μL of methanol and homogenized at 1,100 rpm for 120 s.After centrifugation at 2,300 × g for 5 min at 4°C, 400 μL of the aqueous phase was filtered using a 5-kDa cut-off filter (Ultra Free MC PL-HCC, HMT) by centrifugation at 9,100 × g for 120 min at 4°C.The filtrates were desiccated and resuspended in 50 μL of Milli-Q water (MQ) for CE-TOFMS analysis.
To investigate lipids such as FA, liquid chromatography time-of-flight mass spectrometry (LC-TOFMS) was used.Five hundred microliters of 0.1% formic acid and acetonitrile with 10 μM internal standard and 167 μL of MQ was added to each cheese sample (35-40 mg), and the samples were homogenized.After centrifugation at 2,300 × g for 5 min at 4°C, supernatants were collected.The precipitates were treated with 500 μL of 0.1% formic acid and acetonitrile, homogenized, and centrifuged as described above.The 2 collected supernatants were then mixed.The mixed supernatants were filtered through a 3-kDa cut-off filter by centrifugation at 9,100 × g, 120 min, 4°C (NANOSEP 3K Omega, Pall).Phospholipids were eliminated from the filtrates using a hybrid SPE phospholipid 55261-U filter (Supelco).The filtrates were desiccated and then dissolved in 200 μL of 50% (vol/vol) isopropanol/MQ for LC-TOFMS.
The prepared samples were analyzed using an Agilent 1200 series RRLC system SL equipped with an ODS column (2 × 50 mm, 2 μm) with Agilent LC/ MSD TOF (Agilent Technologies), as described previously (Ohnishi et al., 2017).Under cationic analysis conditions (MS ionization mode: ESI positive), LC/ MSD TOF was conducted as follows: column temperature, 40°C; mobile phase A, H 2 O/0.1% HCOOH; mobile phase B, isopropanol: acetonitrile: H 2 O (65:30:5)/0.1% HCOOH, 2 mM HCOONH 4 ; flow rate, 0.3 mL/min; run time, 20 min; post-time, 7.5 min; gradient condition, 1% mobile phase B (0-0.5 min), 1-100% mobile phase B (0.5-13.5 min), 100% mobile phase B (13.5-20 min); MS nebulizer pressure, 276 kPa; MS dry gas flow, 10 L/min; MS dry gas temperature, 350°C; MS capillary voltage, 4,000 V; MS scan range, m/z 100-1,700; sample injection, 1 μL.Under anionic analysis conditions (MS ionization mode: ESI negative), MS capillary voltage was changed to 3,500 V.The CE-TOFMS and LC-TOFMS data were analyzed using MasterHands ver.2.17.1.11software (Keio University, Tsuruoka, Japan).Metabolites were identified based on m/z and migration time (CE-TOFMS) or retention time (LC-TOFMS).The relative area was calculated as follows: relative area = each peak area/(peak area of internal control × sample volume).For calculating the ratio (sake lees/control) in comparative analyses, the relative area for the cheese supplemented with sake lees was divided by the equivalent area for the control cheese.Differences between the cheese supplemented with sake lees (n = 3) and the control cheese (n = 3) were assessed using the 2-tailed Welch's t-test.P-values of < 0.05 were considered to be significant.Results for the relative area were expressed as mean ± standard deviation.Principal component analysis was conducted using the house statistical software by the HMT.

Cheesemaking
The supplementation of sake lees used in this study was quite different from the koji cheese-making method used by Nakanishi and Nakazawa (1965).The optimal growth pH of A. oryzae is known to be pH 5 to 6 (Gomi, 2014).Sake lees supplementation can accelerate the decrease in pH of the curd (Figure 1B).Thus, to prevent lowering of pH, curd washing was performed by replacing 75% (vol/vol) of the whey with water, to eliminate lactose, leading to lactic acid production by LAB.Koji cheese was ripened at 30°C to 35°C because the optimal temperature of A. oryzae is 35°C to 38°C, and its protease production is strong at 30°C (Narahara et al., 1982;Yamashita, 2021).In addition, the ripening temperature of koji cheese was changed from 30°C to 10°C in phases to prevent, as much as possible, the growth of Escherichia coli or other contaminants.Although further study may be needed to improve the method used to make koji cheese with sake lees, A. oryzae grew successfully on the cheese curd, and pathogenic bacteria such as Listeria and E. coli were not detected in this study (data not shown).

LAB Counts and pH in the Cheeses
The LAB counts in cheese supplemented with sake lees (S0 and S30) were higher than those in the control cheese both before and after ripening (C0 and C30; Figure 1).No microorganisms, such as yeast and LAB, were detected in the sake lees (data not shown).Sake lees promotes the growth of LAB because of its nutritional value LAB, such as vitamins and AA (Izu et al., 2019).Reflecting this result, the pH of S0 was lower than that of C0.Metabolome analysis results for S0 showed a lactate level 1.1-fold that in C0 and a succinic acid level 1.2-fold that in C0, although the differences were not significant (data not shown).Several organic acids may have been responsible for the decrease in pH.Interestingly, after the ripening period, the pH of S30 was higher than that of C30.Fungi can consume lactate and produce ammonia (Suzuki et al., 2021).Although there was little difference between the amount of lactate in C30 and S30 (the lactate level in S30 was 0.9-fold that in C30; data not shown), ammonia production by A. oryzae may have caused the change in pH in the cheese.Further study would be needed to clarify why changes occur in pH following sake lees supplementation.

Cheese Color
Figure 2 presents the results for the color testing of the inner part of the cheese (the rind had been removed).The mean L* values for C0, S0, C30, and S30 were 90.2, 89.7, 84.5, and 82.6, respectively.The mean a* and b* values for C0, S0, C30, and S30 were −1.44, −1.31, −1.46, and −0.16, and 6.76, 7.94, 10.35, and 11.71, respectively.The L* and b* values in the cheese before ripening were higher and lower, respectively, than those in the cheese ripened for 30 d.These color differences could have been affected by the ripening period.The value of a* in S30 was significantly higher than the values for the other cheese samples (P < 0.05).Generally, a decrease in whiteness (L*) is observed as the cheese ripens because casein or other components, such as lipids, are degraded (Milovanovic et al., 2020).The values for a* and b* are considered to be correlated with the ripening period, with ripening increasing b* but not affecting a* in the interior of the cheese (Álvarez and Fresno, 2020).However, the aminocarbonyl reaction between carbohydrate and protein in both sake lees and cheese could have affected the color of the cheese.

Metabolome Analysis
Following CE-TOFMS and LC-TOFMS analysis, 279 metabolites (cationic: 173, anionic: 106) and 147 metabolites (positive mode: 85, negative mode: 62) were identified.Principal component analysis clearly showed differences between the cheeses (Figure 3).The clear class separations indicated that the compositional profile differed in association with sake lees supplementation and with the ripening period.The distance between the samples with or without sake lees was large for the samples after ripening, whereas for those before ripening, the separation was less, suggesting that ripening duration contributed to the discrimination.Table 1 shows the levels of metabolites found in the cheeses; the lower rows show which metabolites were present in S30 cheese at levels (P < 0.05) at least double those in C30, whereas the upper rows show metabolites not detected in C30 cheese (the data for the relative area of C30 and S30 and the ratio S30/C30 are highlighted with a gray background).Concerning metabolites not detected in C30 but detected in S30, monosaccharides (N-acetylgalactosamine or N-acetylmannosamine or Nacetylglucosamine) were detected only in S30 (although these metabolites were also detected in both C0 and S0).There was no significant difference in monosaccharide levels between S0 and S30.These metabolites,  (Robinson, 2019), were evidently retained in the koji cheese supplemented with sake lees, although the reason for this is unknown.
Metabolites such as phytosphingosine, AEA [acylethanolamide (18:3)], AC [acylcarnitine (17:0)], linoleyl ethanolamide, curcumin, ethyl arachidonate, and FA(24:0) were detected in S0 and S30.These metabolites (excluding ethyl arachidonate) were detected in and likely derived from sake lees.Other metabolites such as 5-hydroxyindoleacetic acid, N-acetylmethionine,  betaine aldehyde, histamine, and N 2 -acetylaminoadipic acid were detected only in S30 (and not detected in sake lees), and a synergistic effect between koji and sake lees may have contributed to their production.Histamine is associated with allergic reactions.The concentration of histamine in S30 was 0.480 mg/100 g (by HPLC analysis; data not shown).Although the concentration of histamine differs between types of cheese, the range is 0.3 to 250 mg/100 g (Chambers and Staruszkiewicz, 1978).In this study, the amount of histamine found in the koji cheese with sake lees was considered to be lower than in other cheeses.
Among metabolites in S30 whose levels were 2-fold those in C30 (detected in both C30 and S30 cheeses), the amount of sphinganine in S30 was 199-fold that in C30.The amount of this metabolite in S0 was also 129-fold that in C0.Therefore, sphinganine was considered to be transferred from sake lees because the metabolite was detected in sake lees (data not shown).Levels of other metabolites such as linoleic acid, sphinganine (d20:0), sphingosine, and palmitoylethanolamide in S0 and S30 were also higher than those in C0 and C30, respectively.These metabolites were also detected in sake lees and considered to be transferred to the cheese.Levels of the FA(20:3), known as dihomo-gamma-linolenic acid, in S0 and S30 were higher than those in C0 and C30, respectively.The FA(20:3) is produced by A. oryzae, a phenomenon that has been investigated because of its antiinflammatory and antiproliferative effects (Tamano et al., 2020).Sake lees supplementation may have enhanced FA(20:3) production by A. oryzae in the cheese.The amount of zeaxanthin in S30 was also 2-fold that in C30.A mechanism for the increase in zeaxanthin is not clear; a relevant biosynthesis pathway in A. oryzae could not be found in the KEGG database (https: / / www .genome.jp/pathway/ aor00906).
Table 2 shows metabolites detected at reduced levels in the koji cheese supplemented with sake lees.Because       , each group: control cheese without sake lees (C0), cheese supplemented with sake lees (S0), C0 ripened for 30 d (C30), S0 ripened for 30 d (S30).ND = not detected; NA = not available (not detected in one of cheeses for comparison). 2 Compound name (metabolite library maintained by Human Metabolome Technologies, Tsuruoka, Japan) based on mass-to-charge ratio (m/z) and migration time (MT).RT = room temperature.UDP = uridine diphosphate. 3 The relative area = each peak area/(peak area of internal control × sample volume). 4 The ratio means [the relative area of the cheese supplemented with sake lees (S0 and S30)/the relative area of the control cheese (C0 and C30)] or [the relative area of the ripened cheese (C30 and S30)/the relative area of the preripening cheese (C0 and S0)].
*P < 0.05, **P < 0.01, ***P < 0.001 by Welch t-test.Compound name (metabolite library maintained by Human Metabolome Technologies, Tsuruoka, Japan) based on mass-to-charge ratio (m/z) and migration time (MT).RT = room temperature.GABA = γ-aminobutyric acid. 3 The relative area = each peak area/(peak area of internal control × sample volume). 4 The ratio means [the relative area of the cheese supplemented with sake lees (S0 and S30)/the relative area of the control cheese (C0 and C30)] or [the relative area of the ripened cheese (C30 and S30)/the relative area of the preripening cheese (C0 and S0)].
*P < 0.05, **P < 0.01, ***P < 0.001 by Welch t-test.the amounts of uridine diphosphate (UDP)-glucose (or UDP-galactose), UDP-N-acetylgalactosamine (or UDP-N-acetylglucosamine), and glucuronic acid (or galacturonic acid) in S30 were lower than those in C30, nucleotide sugar metabolism seems to be influenced by sake lees.Additionally, overall, levels of peptides such as carnosine, Ile-Pro-Pro, Arg-Glu, Ser-Glu, β-Ala-Lys, and Glu-Glu in S30 were lower than those in C30, even though peptides can be produced by LAB after degradation of milk proteins (Hagi et al., 2016(Hagi et al., , 2019)).The diversity of peptides might be reduced by supplementation of sake lees because there was an increase in the number of LAB in cheese supplemented with sake lees (Figure 1).Additionally, carbon flux, such as through monosaccharide (Table 1: N-acetylgalactosamine or N-acetylmannosamine or N-acetylglucosamine) and nucleotide sugar metabolism (Table 2), caused by sake lees might also have affected peptidase activities because peptidase in koji is influenced by carbon flux (Zhao et al., 2019).
Free AA can be used as an index of ripening in cheese (Ji et al., 2004).In this study, metabolome analysis showed no significant differences in levels of any of the AA studied between C30 and S30 (Table 3), indicating that sake lees supplementation had no significant effect on AA production in cheese.Before ripening, the amounts of methionine and tryptophan in S0 were higher than those in C30.These AA are derived from sake lees.Although sake lees contain other AA, there were only small differences in levels between samples of C0 and S0.Other AA found in sake lees may be present only in small amounts or are metabolized by LAB during the mold-drain process, activating LAB fermentation resulting in an increase in LAB number and subsequent decrease in pH in cheese before ripening.In contrast to AA, levels of 4 γ-glutamyl peptides (γ-Glu-Gly, γ-Glu-Ala, γ-Glu-Ile or γ-Glu-Leu, and γ-Glu-Phe) were increased in cheese supplemented with sake lees (Table 1).Gamma-glutamyl peptides are known as kokumi-active glutamyl peptides, and they are found in cheeses cultured with P. roqueforti (Toelstede and Hofmann, 2009), where they are important components of the taste of the cheese (kokumi means "rich taste" in Japanese).Table 4 shows levels of all of the γ-glutamyl peptides detected in the cheese in this study and their ratio (S30/C30); levels of 12 of the 16 γ-glutamyl peptides detected were significantly higher in S30 than in C30 (P < 0.05).Gamma-glutamyl-Leu and γ-glutamyl-Val are known for their superior umamienhancing effect (Yang et al., 2021).Supplementation with sake lees could improve the taste of koji cheese by increasing levels of various γ-glutamyl peptides, including γ-glutamyl-Leu and γ-glutamyl-Val.In sake lees, all of the γ-glutamyl peptides were detected (data not shown); however, the levels were considered to be low because these peptides were not detected in the cheeses before ripening (C0 and S0).The γ-glutamyl peptides were produced by the koji during the ripening process, as demonstrated by the increased levels of these peptides in both C30 and S30.It is known that A. oryzae possesses glutaminase and can thus produce γ-glutamyl peptides (Yang et al., 2017).Reportedly, sake lees can enhance the production of AA and peptides in sausages (Mikami et al., 2020).Our study has shown that sake lees enhance the production of γ-glutamyl peptides in cheese.Supplementation with sake lees could enhance proteinase and glutaminase degradation of proteins.

CONCLUSIONS
This study provides further insights into cheese ripened by koji fungi (A.oryzae).During koji cheese production, supplementation with sake lees increased the number of LAB, which reduced pH and changed the color of the cheese.Metabolome analysis revealed changes in koji cheese carbohydrate levels caused by sake lees, with decreased amounts of nucleotide sugars, such as UDP-glucose (or UDP-galactose), UDP-Nacetylgalactosamine (or UDP-N-acetylglucosamine), and glucuronic acid, being detected.Interestingly, supplementation with sake lees enhanced the production of γ-glutamyl peptides that provide an umami-enhancing effect.Information from this study will contribute to further development in the field of cheese production.

Figure 1 .
Figure 1.Number of lactic acid bacteria (LAB; A) and pH (B) in cheese.C0 and S0: preripened cheeses without and with sake lees supplementation.C30 and S30: the same cheeses ripened for 30 d. Mean ± SD of triplicates for each group.Asterisks (*) indicate significant difference (P < 0.05).

Figure 2 .
Figure 2. The effect of sake lees on the color of the cheese.C0 and S0: preripened cheeses without and with sake lees supplementation.C30 and S30: the same cheeses ripened for 30 d. Mean ± SD of triplicates for each group.Different lowercase letters indicate significant differences (P < 0.05).The color measured according to the CIELAB color system presented as lightness (L*; A), redness (a*; B), and yellowness (b*; C).

Figure 3 .
Figure 3. Principal component (PC) analysis for cheeses in the study.C0 and S0: preripened cheeses without and with sake lees supplementation.C30 and S30: the same cheeses ripened for 30 d. Data points for each triplicate in each group are shown.

2
Compound name (metabolite library maintained by Human Metabolome Technologies, Tsuruoka, Japan) based on mass-to-charge ratio (m/z) and migration time (MT).RT = room temperature.AEA = acylethanolamide; AC = acylcarnitine; FA = fatty acid.3Therelative area = each peak area/(peak area of internal control × sample volume).4Theratio means [the relative area of the cheese supplemented with sake lees (S0 and S30)/the relative area of the control cheese (C0 and C30)] or [the relative area of the ripened cheese (C30 and S30)/the relative area of the preripening cheese (C0 and S0)].*P < 0.05, **P < 0.01, ***P < 0.001 by Welch t-test.
each group: control cheese without sake lees (C0), cheese supplemented with sake lees (S0), C0 ripened for 30 d (C30), S0 ripened for 30 d (S30).ND = not detected; NA = not available (not detected in one of cheeses for comparison).2Compound name (metabolite library maintained by Human Metabolome Technologies, Tsuruoka, Japan) based on mass-to-charge ratio (m/z) and migration time (MT).RT = room temperature.3Therelative area = each peak area/(peak area of internal control × sample volume).4Theratio means [the relative area of the cheese supplemented with sake lees (S0 and S30)/the relative area of the control cheese (C0 and C30)] or [the relative area of the ripened cheese (C30 and S30)/the relative area of the preripening cheese (C0 and S0)].*P < 0.05, **P < 0.01, ***P < 0.001 by Welch t-test.
Hagi et al.: KOJI CHEESE SUPPLEMENTED WITH SAKE LEES derived from milk

Table 1 .
Hagi et al.: KOJI CHEESE SUPPLEMENTED WITH SAKE LEES Metabolites detected at higher levels in koji cheese supplemented with sake lees than that in the unsupplemented cheese 1

Table 2 .
Hagi et al.: KOJI CHEESE SUPPLEMENTED WITH SAKE LEES Metabolites detected at lower levels in koji cheese supplemented with sake lees than that in unsupplemented cheese 1

Table 3 .
Hagi et al.: KOJI CHEESE SUPPLEMENTED WITH SAKE LEES Differences in AA content between preripened (control) cheese and cheese supplemented with sake lees 1 2

Table 4 .
Hagi et al.: KOJI CHEESE SUPPLEMENTED WITH SAKE LEES Gamma-glutamyl peptides detected in koji cheese 1 Compound name 2