Article pubs.acs.org/JAFC
Metabolic Profiling Approach To Explore Compounds Related to the Umami Intensity of Soy Sauce Kazuki Shiga,†,‡ Shinya Yamamoto,† Ayako Nakajima,‡ Yukako Kodama,‡ Miho Imamura,‡ Tsuneo Sato,‡ Riichiro Uchida,‡ Akio Obata,‡ Takeshi Bamba,† and Eiichiro Fukusaki*,† †
Department of Biotechnology, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka 565-0871, Japan Research and Development Division, Kikkoman Corporation, 399 Noda, Noda City, Chiba 278-0037, Japan
‡
S Supporting Information *
ABSTRACT: A metabolic profiling approach was employed to explore the compounds that affect the intensity of umami taste in soy sauce. Twenty-five kinds of soy sauces were analyzed using GC-MS and LC-MS, wherein measurement data for 427 compounds were obtained. The umami taste intensity of each soy sauce sample was also quantitated by sensory evaluation and a projection to latent structure (PLS) regression analysis was conducted using the compounds’ measurements and umami taste intensity data. Variable importance for the projection (VIP) value obtained via PLS was used for the estimation of the relevance for umami taste intensity. N-(1-Deoxyfructos-1-yl)glutamic acid (Fru-Glu) had the highest VIP value, and addition of Fru-Glu to soy sauce increased umami taste intensity better than glutamic acid at the same concentration as confirmed by sensory evaluation. This study showed that the combination of metabolic profiling approach and sensory evaluation can be used effectively to determine compounds related to taste. KEYWORDS: soy sauce, metabolic profiling, umami, GC-MS, LC-MS
■
sauce is around 1.2% (w/v).6 This suggests that MSG, which has a significant contribution in the umami taste, does not necessarily play a large role in distinguishing umami tastes among soy sauce samples and that other umami compounds have a significant effect on the difference between umami taste intensities. Therefore, it is worthwhile to focus on other compounds aside from MSG in the study of the relationship between compounds present in soy sauce and umami taste. Many investigations reported that compounds in soy sauce such as aspartic acid, phenylalanine, peptides including pyroglutamyl peptide, and amadori compounds are the compounds related to umami taste.7−9 Another study suggested that low-molecular-weight compounds in soy sauce contribute to the umami taste.10 The taste characteristics of these compounds were evaluated on model experiments using either solely aqueous solutions or MSG mixture solution; however, the results of model experiments did not always correspond to the results of complex food systems considering the interactions of taste−taste or among taste compounds.11 Thus, it is still unclear what the compounds that affect the umami taste intensity in actual soy sauce are. In the case of exploratory approach for umami compounds of soy sauce, the sensory-guided fractionation technique of soy sauce using preparative chromatographic separation is often used.9,10 This technique is useful for the investigation of the taste contributants in a few samples. On the other hand, because soy sauce products have a wide variety, it is preferred
INTRODUCTION Soy sauce is a traditional brewed seasoning in Eastern Asia, and it has been consumed widely in the world due to its characteristic taste and flavor. Even though soy sauce is explained as a seasoning, which is produced by the fermentation of soybeans, soy sauce has many varieties due to differences in the country of origin, region, ingredients, production process, manufacturers, etc. For example, the starting material of koikuchi soy sauce, which is mainly consumed in Japan, is equal amounts of soybeans and wheat.1 On the other hand, common Chinese soy sauce is made from only soybeans or sometimes mixed with a small amount of wheat.2 These varieties of soy sauce products generate differences in the taste characteristics or intensity of each soy sauce product. Even though there is variety among soy sauce products, soy sauce products commonly have five basic tastes; sweet, bitter, salty, sour, and umami. Especially, umami is one of the most important criteria for evaluating the quality of soy sauce. Hence, it is significant to clearly explain the relationship between the umami taste and compounds present in the soy sauce. There are many studies about umami compounds in soy sauce in which some suggested that the key compound of the umami taste of soy sauce is monosodium L-glutamate (MSG).3−5 However, it is impossible to completely attribute the umami taste of soy sauce to MSG.3−5 Moreover, the discrimination threshold of MSG in Japanese koikuchi soy sauce is 1.9% (w/v), even though the detection threshold of MSG in aqueous solution is 0.026% (w/v).6 This means that the addition of >2 times of the concentration of MSG in soy sauce is necessary to recognize the difference of the umami intensity of soy sauce, considering the concentration of MSG in Japanese koikuchi soy © 2014 American Chemical Society
Received: Revised: Accepted: Published: 7317
March 9, 2014 June 16, 2014 June 21, 2014 June 21, 2014 dx.doi.org/10.1021/jf501173r | J. Agric. Food Chem. 2014, 62, 7317−7322
Journal of Agricultural and Food Chemistry
Article
resin, Wako Pure Chemical Industries, Osaka, Japan) and ODS-type resin (Daisogel SP-120-40/60-ODS-B, DAISO Co., Ltd., Osaka, Japan). The purity of Fru-Glu (>90%) was confirmed by LC-MS using the same conditions as given below. Sample Preparation of Soy Sauce for GC-MS Analysis. A previously reported method of sample preparation and derivatization for GC-MS analysis18 was used with modifications. Soy sauce samples were diluted 10 times, and 20 μL of the diluted sample in a microtube was extracted with 1000 μL of MeOH/H2O/CHCl3 (5:2:2, v/v/v). Sixty microliters of 0.2 mg/mL ribitol was added to the mixture. The sample was centrifuged at 16000g, for 3 min at 4 °C, and then 900 μL of the supernatant was transferred to another microtube. Four hundred microliters of distilled water was added to the sample before mixing. Following centrifugation (16000g, 3 min at 4 °C), 900 μL of supernatant was transferred to another microtube and capped. The cap was subsequently pierced and the extract was evaporated, to remove methanol, in a centrifuge vacuum concentrator at room temperature for approximately 2 h. After evaporation, the extract was freeze-dried at room temperature overnight. Sample Derivatization for GC-MS Analysis. For derivatization, 100 μL of methoxyamine hydrochloride in pyridine (20 mg/mL) was added to the above freeze-dried sample, and the mixture was incubated in a Thermomixer comfort (Eppendorf Ltd., Hamburg, Germany) at 30 °C for 90 min to induce the methoxylation reaction. Fifty microliters of N-methyl-N-(trimethylsilyl)trifluoroacetamide, the second derivatization agent, was added, and the mixture was incubated at 37 °C for 90 min to induce the silylation reaction. The samples were then used for GC-MS analysis. One microliter of sample was injected in split mode (25:1, v/v). GC-MS Analysis. A previously reported GC-MS condition18 was used with modifications. The GC-MS used in this study was a Shimadzu GCMSQP2010 Ultra (Shimadzu, Kyoto, Japan) equipped with a 30 m × 0.25 mm i.d. fused silica capillary column coated with 0.25 μm CP-SIL 8 CB low bleed/MS (Agilent Technologies, Palo Alto, CA, USA). The injection temperature was 230 °C. The carrier gas was helium at a linear velocity of 39 cm/s. The column temperature was held at 80 °C for 2 min isothermally and then raised at 15 °C/min to 330 °C and held for 6 min. The MS conditions were as follows: transfer line temperature, 250 °C; ion source temperature, 200 °C. Ions were generated at 70 eV with electron ionization and were recorded at 20 scan/s over the mass range m/z 85−500. Each sample was measured three times. A standard alkane mixture (C8− C40) was injected at the beginning and end of the analysis for tentative identification. Sample Preparation of Soy Sauce for LC-MS Analysis. The sodium chloride (NaCl) level in soy sauce was reduced to below 1.0% by electrodialysis with a Micro Acilyzer S1 (Asahi Chemical Industry, Tokyo, Japan). One hundred microliters of soy sauce sample and 200 μL of methanol were mixed, and 10 μL of the mixed sample was diluted 10 times with distilled water. The sample was filtered before it was placed in a glass vial. LC-MS Analysis. LC experiments were conducted using an HPLC system (Shimadzu, Japan) consisting of an LC-20AB binary pump, an SIL-20AC autosampler, and a CTO-20AC column oven. Chromatographic separation was achieved using a reversed-phase column (Discovery HS-F5 column, 150 mm × 2.1 mm, 3.5 μm, Sigma-Aldrich, Milwaukee, WI, USA) with a gradient elution with a mobile phase composed of eluents A (0.1% acetic acid in water, v/v) and B (0.1% acetic acid in acetonitrile, v/v). The mobile phase was consecutively programmed as follows: an isocratic elution of 2% B for the first 3 min, followed by linear gradient elutions of 2−50% B from 3 to 16 min and 50−100% B from 16 to 18 min. After the solvent composition of 100% B had been held from 18 to 24 min and then changed to 2% B for the next 10 min, the column was returned to its starting conditions. A 1 μL aliquot of the sample solution was injected onto the column. The flow rate was 0.2 mL/min, and the column temperature was maintained at 40 °C throughout the analysis. The compounds were detected by Shimadzu ion trap/time-of-flight hybrid mass spectrometry (IT-TOF/ MS) equipped with an electrospray ionization source (ESI). The MS conditions were as follows: positive ion mode; electrospray voltage, 1.7
that several kinds of soy sauce are assessed. However, this technique is expensive and time-consuming due to the preparation of a huge amount of fraction of soy sauce. Thus, it is difficult to use this technique for several soy sauce samples. Meanwhile, metabolomics concerning as many metabolites as possible has become an important tool over a wide research area. Specifically, metabolic or compound profiling based on metabolomics technology has recently been receiving much attention for application in the research of the relationship between compounds and food quality.12,13 Research on metabolic profiling and its relationship to the taste of food has been reported before such as in green tea,14 tomato,15 cheese,16 sake,17 and soy sauce.18 In a previous study18 about profiling between the compounds and taste of soy sauce, hydrophilic compounds were measured by gas chromatography−mass spectrometry (GC-MS), and its correlation with sensory attribute was assessed. The study focused on the construction of a prediction model of sensory attributes by metabolic profiling that was able to predict the umami taste intensity using metabolite data obtained by GCMS. In addition, it was reported that glutamic acid, glucose, and sucrose had strong correlation with the umami taste intensity and were suggested as the significant compounds for umami taste. However, the study did not confirm the effect of these compounds on the umami taste of actual soy sauce. Moreover, the compounds’ data measured by GC-MS were not likely enough for the aim of the study due to the limitation of measurable compounds by GC-MS. The aim of the present study was to explore the compounds that affect the umami taste intensity in actual soy sauce by using metabolic profiling and sensory evaluation. We employed two analytical platforms, GC-MS and liquid chromatography−mass spectrometry (LC-MS), for the measurement of compounds. For selection of the candidate compounds related to the umami taste intensity, projection to latent structure (PLS) regression analysis (also known as partial least-squares) was conducted with both the metabolite data and the umami taste intensity data. Additionally, the effect of the candidate compound on the umami taste intensity in soy sauce by sensory evaluation was assessed. On the basis of a metabolic profiling approach, we show that the compound found to be statistically significant to umami taste intensity indeed affects the umami taste of soy sauce.
■
MATERIALS AND METHODS
Soy Sauce. Twenty-five soy sauce samples were used in this study. The countries of origin of the samples were China (samples 1−6 and 20−22), the United States (samples 7−9), and Japan (samples 10−19 and 23−25). The samples were preserved in a freezer at −30 °C prior to chemical analysis and sensory evaluation. Chemicals. All chemicals used in this study were of analytical grade. Ribitol was used as internal standard, and pyridine was used as solvent; both were purchased from Wako Pure Chemical Industries (Osaka, Japan). The derivatization reagents methoxyamine hydrochloride and N-methyl-N-(trimethylsilyl)trifluoroacetamide (MSTFA) were purchased from Sigma-Aldrich (Milwaukee, WI, USA) and GL Science, Inc. (Tokyo, Japan), respectively. N-(1-Deoxyfructos-1-yl)glutamic acid (Fru-Glu) for compound identification and sensory evaluation was synthesized from food grade glucose and glutamic acid following the literature.19 Briefly, a mixture of D-glucose (Sanei Sucrochemical Co., Ltd., Aichi, Japan) and L-glutamic acid in a molar ratio of 40:1 in 50 mM phosphate buffer was heated at 60 °C for 3 days under reflux with vigorous stirring. The reaction mixture was purified by passage through a column of cationexchange resin (DOWEX 50Wx8 200−400 mesh (H) cation exchange 7318
dx.doi.org/10.1021/jf501173r | J. Agric. Food Chem. 2014, 62, 7317−7322
Journal of Agricultural and Food Chemistry
Article
kB; CDL temperature, 200 °C; block heater temperature, 200 °C; nebulizing gas (N2), 1.5 L/min; drying gas (N2) pressure, 0.1 MPa. Mass spectra were acquired in the range of m/z 100−1000, and ion accumulation time was 200 ms. The calibration of mass accuracy was conducted every 24 h with standard. For quality control, the mixture of all samples was measured every 24 h during the measurement period. Each sample was measured three times. In MS/MS analysis, MS/MS spectra were acquired in the m/z range of 80−500 under the following conditions: precursor ion isolation in the m/z range of 100−1000 (width, 3 Da); ion accumulation time, 50 ms; tolerance m/z, 0.05; repeat = 2; and CID energy, 50%. In quantitative analysis, we used synthesized Fru-Glu for external standard and the LC-MS system under the same conditions as above. Data Processing. A previously reported method for data processing concluding tentative peak identification18 was used with modifications. MS data obtained by GC-MS were exported in the netCDF format. Peak detection and alignment were performed by MetAlign software (Wageningen UR, The Netherlands; freely available at http://www.pri.wur.nl/UK/products/MetAlign/). 20 The resulting data were exported in the CSV-format file. Then, the CSV-format file was imported by the AIoutput software21 for tentative peak identification and construction of a data matrix. In tentative peak identification, retention indices of the eluted compounds were calculated on the basis of the standard alkane mixture. Then, by comparing the retention indices and unique mass spectra with our inhouse reference library, tentative identifications were obtained. In the construction of the data matrix, spectra were normalized manually by adjusting the peak intensity against the ribitol internal standard before an organized data matrix was generated. The peak detection and alignment of MS data obtained by LC-MS were performed by Profile solution version 1.0 (Shimadzu). The peak integration parameters were set as default, and the peak alignment parameters were as follows: ion m/z tolerance, 5 mDa; retention time tolerance, 0.5 min; peak intensity threshold, 5000. The resulting data were exported in the CSV-format file. To reduce the risk of ghost peaks, peaks having maximum intensity and minimum relative standard deviation among samples of >300,000 and 1” rule is generally used as the criterion for important variable selection.24
■
RESULTS AND DISCUSSION Metabolite Data Obtained by GC-MS and LC-MS. We employed two analytical platforms, GC-MS and LC-MS, for the measurement of compounds in 25 samples of soy sauce. GCMS was used to measure low hydrophilic nonvolatiles such as sugar or amino acid with sample derivatization, whereas LC-MS was expected to detect the compounds that are difficult to volatilize with derivatization such as peptides and Maillard compounds. Using the two platforms made it possible to detect a broad range of physical and chemical characteristics of compounds in soy sauce. From the 122 peaks detected in the GC-MS analysis, 68 peaks were tentatively identified by comparing the retention indices and unique mass spectra with our in-house reference library using AIoutput 2.0. The tentatively identified compounds included 26 kinds of amino acids, 23 kinds of sugars or sugar alcohols, 8 kinds of organic acids, 4 kinds of amines, and 7 kinds of other compounds (Supporting Information Table S1). On the other hand, 305 peaks were detected with the positive mode of LC-MS. The m/ z range of these peaks was from 110.008 to 671.200. These results suggested that a wide range of compounds were detected by GC-MS and LC-MS analyses. We used the data obtained from LC-MS for explanatory variables of PLS regression analysis before peak identification because identification of all peaks was difficult using only accurate mass. Then, the selected peaks based on VIP values obtained via PLS (described in the succeeding section) were identified by comparison with the MS/MS spectrum and retention time of authentic standard. Umami Intensity of Soy Sauce by Sensory Evaluation. The umami taste intensity of each soy sauce sample was obtained using sensory evaluation (Figure 1). The highest umami taste intensity was 4.01, and the lowest was 2.79. For evaluation of the significant differences among samples, we conducted one-way ANOVA. The F value was 1.53, and p value was 1 were considered to be significant.24 These compounds in Table 1 have VIP values >1; thus, these compounds are significant explanatory variables for the prediction of the response variable. In the following discussion, we focused on four compounds that had VIP values larger than that of glutamic acid. In Table 1, both Fru-Glu and inosinic acid were identified using information on both the retention time and MS/MS spectra of authentic standards obtained by LC-MS. Fru-Glu was detected as m/z 310.113 ion and a retention time of 2.20 min. Figure 3 shows the MS/MS spectrum of Fru-Glu obtained by IT-TOF/MS. The m/z 148.06 product ion contained in the mass spectrum corresponded to glutamic acid H+ adduct. The neutral loss from precursor ion (m/z 310.113) is m/z 162.05 and corresponds to the molecular weight of glucose without one water molecule. Inosinic acid was detected as the m/z
Figure 2. Relationships between predicted and measured umami taste intensity based on projection to latent structure (PLS) model. The numbers of explanatory variables and response variables were 427 and 1, respectively. PLS component was 4. R2 and Q2 were 0.914 and 0.877, respectively.
(proportion of predicted variance), which are the diagnostic indices of model quality, were 0.914 and 0.877, respectively. Previously, it has been reported that the accuracy of the prediction model of PLS constructed with only GC-MS is enough.18 When GC-MS data obtained in this study were used for PLS, R2 and Q2 were 0.899 and 0.821, respectively. On the other hand, R2 and Q2 from the LC-MS data were 0.909 and
Figure 3. MS/MS spectrum of N-(1-deoxyfructos-1-yl)glutamic acid obtained in the positive ionization mode. Triangle symbols indicate m/ z 148.061 corresponding to glutamic acid H+ adduct. 7320
dx.doi.org/10.1021/jf501173r | J. Agric. Food Chem. 2014, 62, 7317−7322
Journal of Agricultural and Food Chemistry
Article
349.053 ion and retention time of 2.07 min by LC-MS via MS/ MS analysis (data not shown). The compound with the largest VIP value was Fru-Glu. It was reported that the umami taste intensity of Fru-Glu was similar to that of MSG in water or NaCl solution,27 and FruGlu was detected in the wheat gluten hydrolysate.28 However, there has been no report of the umami of Fru-Glu and its taste property in soy sauce. The compound with the second largest VIP detected by GCMS could not be identified using our in-house library. The third largest VIP value was 2.23 for inosinic acid, known as a umami enhancer of MSG.29 In naturally fermented soy sauce, inosinic acid is not detected because the phosphatase enzymes made by kouji mold degrade the inosinic acid in natural soy sauce.30 Also, some soy sauce products have inosinic acid added as a food additive for umami enhancement after fermentation.31 The compound with the fourth largest VIP detected by GCMS was 4-aminobutyric acid. 4-Aminobutyric acid, which has a VIP value was close to that of glutamic acid, has been reported to give a weak sour taste and a mouth-drying sensation in aqueous solution.32 It is unclear whether 4-aminobutyric acid has an effect on the umami taste intensity of soy sauce. This metabolic profiling approach helps in the selection of candidate compounds from a huge number of variables. However, this approach could not substantiate the effect for the umami taste intensity of “actual” soy sauce. We therefore evaluated the effect of Fru-Glu addition, the compound with the highest VIP value, on the umami taste intensity of soy sauce. Effect of Fru-Glu on the Umami Taste Intensity of Soy Sauce. Using sensory evaluation, we assessed the effect of FruGlu addition, the compound with the largest VIP value, on the umami taste intensity of soy sauce. In sensory evaluation, panelists compared the control sample (Japanese koikuchi soy sauce (sample 10) without additional Fru-Glu) with the test sample containing the same soy sauce with additional Fru-Glu. Before sensory analysis, the concentration of Fru-Glu was measured in all samples. We found that the concentration of Fru-Glu in sample 10 was 0.7 mM, and the highest concentration among the samples was 5.5 mM (Figure 4). The additional Fru-Glu concentrations were 5 and 10 mM. The same evaluation method was conducted using glutamic acid. The result of sensory evaluation is shown in Figure 5. The
Figure 5. Results of the sensory evaluation of N-(1-deoxyfructos-1yl)glutamic acid (Fru-Glu) in soy sauce. The umami taste intensity was evaluated by comparison between soy sauce with the addition of FruGlu and without the addition (control) of Fru-Glu. Sample 10 was used for the evaluation. The 9-point relative categorical scale (comparing control sample −4, “very weak”, to 4, “very strong”) was used. Intensity was evaluated by 10 selected panelists. The same evaluation method was used for the addition of glutamic acid (Glu). Values are the mean ± standard error. (∗) p < 0.05 (sample t test against 0).
umami taste intensity tended to increase with higher Fru-Glu concentration, and there was a significant difference between the sample with added 10 mM Fru-Glu and control sample (one-sample t test p < 0.05). On the other hand, there was no significant difference in the umami taste intensity in samples with added 5 and 10 mM glutamic acid. These results indicated that Fru-Glu increased the umami taste intensity of soy sauce at a lower concentration than glutamic acid even though the taste threshold of both compounds in aqueous solutions is the same.24 Consequently, the evaluation of the taste property of compounds in complex systems such as food may differ from that of model experiment due to the effect of taste interactions such as a synergy effect. Furthermore, these results suggested that Fru-Glu has a stronger synergistic effect than glutamic acid in soy sauce. Glutamic acid did not affect the umami taste intensity at this concentration even though it had a high VIP value. Considering the discrimination threshold of MSG in koikuchi soy sauce,6 a concentration >10 mM is needed for increasing the umami taste intensity of soy sauce. We therefore conclude that to enhance umami taste, at low concentration, Fru-Glu is more essential than glutamic acid in soy sauce. To the best our knowledge, there is no report on the relationship between Fru-Glu and umami in soy sauce. In summary, the present study explored the compounds related to the umami taste intensity by a metabolic profiling approach. Using this approach, Fru-Glu was found to increase the umami taste intensity of soy sauce more effectively than glutamic acid, the key umami compound. The advantages of the metabolic profiling approach over traditional methods such as sensory-guided fractionation are as follows: (1) metabolic profiling is able to cover a wide variety samples because it is not necessary to prepare fractions; (2) compounds that are difficult to separate from other compounds can also be evaluated. However, the result of statistical analysis such as PLS does not directly substantiate the contribution of each compound to taste; thus, validation of the effect of compounds on taste using sensory evaluation is necessary. The combination of metabolic profiling and sensory evaluation is a useful tool for exploring the compounds that affect taste and may contribute to the
Figure 4. Concentration of N-(1-deoxyfructos-1-yl)glutamic acid in each soy sauce sample. Values are the mean ± standard deviation in three replications. 7321
dx.doi.org/10.1021/jf501173r | J. Agric. Food Chem. 2014, 62, 7317−7322
Journal of Agricultural and Food Chemistry
Article
by gas chromatography/ mass spectrometry-based hydrophilic. J. Agric. Food Chem. 2007, 55, 231−236. (15) Thissen, U.; Coulier, L.; Overkamp, K. M.; Jetten, J.; van der Werf, B. J. C.; van de Ven, T.; van der Werf, M. J. A proper metabolomics strategy supports efficient food quality improvement: a case study on tomato sensory properties. Food Qual. Pref. 2011, 22, 499−506. (16) Ochi, H.; Naito, H.; Iwatsuki, K.; Bamba, T.; Fukusaki, E. Metabolomics-based component profiling of hard and semi-hard natural cheeses with gas chromatography/time-of-flight-mass spectrometry, and its application to sensory predictive modeling. J. Biosci. Bioeng. 2012, 113 (6), 751−758. (17) Sugimoto, M.; Koseki, T.; Hirayama, A.; Abe, S.; Sano, T.; Tomita, M.; Soga, T. Correlation between sensory evaluation scores of Japanese sake and metabolome profiles. J. Agric. Food Chem. 2010, 58, 374−383. (18) Yamamoto, S.; Bamba, T.; Sano, A.; Kodama, Y.; Imamura, M.; Obata, A.; Fukusaki, E. Metabolite profiling of soy sauce using gas chromatography with time-of-flight mass spectrometry and analysis of correlation with quantitative descriptive analysis. J. Biosci. Bioeng. 2012, 114 (2), 170−175. (19) Roper, H.; Roper, S.; Heyns, K.; Meyer, B. NMR spectroscopy of N-(1-deoxy-D-fructos-1-yl)-L-amino acids (“fructose-amino acids”). Carbohydr. Res. 1983, 116, 183−195. (20) Lommen, A. MetAlign: interface-driven, versatile metabolomics tool for hyphenated full-scan mass spectrometry data preprocessing. Anal. Chem. 2009, 81, 3079−3086. (21) Tsugawa, H.; Tsujimoto, Y.; Arita, M.; Bamba, T.; Fukusaki, E. GC-MS based metabolomics: development of a data mining system for metabolite identification by using soft independent modeling of class analogy (SIMCA). BMC Bioinf. 2011, 12, 131. (22) Eriksson, L.; Johansson, E.; Kettaneh-Wold, N.; Wold, S. Statistical Appendix in Multi- and Megavariate Data Analysis − Principle and Applications; Umetrics AB: Umea, Sweden, 2001; 504 pp. (23) Wold, S.; Johansson, E.; Cocchi, M. 3D QSAR in Drug Design; Theory, Methods, and Applications. ESCOM: Leiden, The Netherlands, 1993; pp 523−550. (24) Afanador, N. L.; Tran, T. N.; Buydens, L. M. C. Use of the bootstrap and permutation methods for a more robust variable importance in the projection metric for partial least squares regression. Anal. Chem. Acta 2013, 768, 49−56. (25) Wold, S.; Sjöström, M.; Eriksson, L. PLS-regression: a basic tool of chemometrics. Intell. Lab. Syst. 2011, 58, 109−130. (26) Fonville, J. M.; Richards, S. E.; Barton, R. H.; Boulange, C. L.; Ebbels, T. M. D.; Nicholson, J. K.; Holmes, E.; Dumas, M. E. The evolution of partial least squares models and related chemometric approaches in metabonomics and metabolic phenotyping. J. Chemom. 2010, 24, 636−649. (27) Beksan, E.; Schieberle, P.; Robert, F.; Blank, I.; Fay, L. B.; Schlichtherle-Cerny, H.; Hofmann, T. Synthesis and sensory characterization of novel umami-tasting glutamate glycoconjugates. J. Agric. Food Chem. 2003, 51, 5428−5436. (28) Schlichtherle-Cerny, H.; Affolter, M.; Cerny, C. Hydrophilic interaction liquid chromatography coupled to electrospray mass spectrometry of small polar compounds in food analysis. Anal. Chem. 2003, 75, 2349−2354. (29) Yamaguchi, S. The synergistic taste effect of monosodium glutamate and disodium 5′-inosinate. J. Food Sci. 1967, 32, 473−478. (30) Kim, J.; Lee, Y. A study of chemical characteristics of soy sauce and mixed soy sauce: chemical characteristics of soy sauce. Eur. Food Res. Technol. 2008, 227, 933−944. (31) Yue, X. Advances in the soy sauce industry in China. J. Ferment. Bioeng. 1990, 70, 434−439. (32) Rotzoll, N.; Dunkel, A.; Hofmann, T. Activity-guided identification of (S)-malic acid 1-O-D-glucopyranoside (morelid) and γ-aminobutyric acid as contributors to umami taste and mouthdrying oral sensation of morel mushrooms (Morchella deliciosa Fr.). J. Agric. Food Chem. 2005, 53, 4149−4156.
quality assessment or improvement of soy sauce and other foods.
■
ASSOCIATED CONTENT
S Supporting Information *
Table S1. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*(E.F.) Phone: +81-6-6879-7424. Fax: +81-6-6879-7424. Email:
[email protected]. Notes
The authors declare no competing financial interest. The study represents a portion of the dissertation submitted by K.S. to Osaka University in partial fulfillment of the requirement for his Ph.D.
■
ABBREVIATIONS USED GC-MS, gas chromatography−mass spectrometry; Fru-Glu, N(1-deoxyfructos-1-yl)glutamic acid; Glu, glutamic acid; LC-MS, liquid chromatography−mass spectrometry; MTSFA, methylN-(trimethylsilyl)trifluoroacetamide; PLS, projection to latent structures; VIP, variable importance for the projection
■
REFERENCES
(1) Kataoka, T. Functional effects of Japanese style fermented soy sauce (Shoyu) and its components. J. Biosci. Bioeng. 2005, 100 (3), 227−234. (2) Sun, S. Y.; Jiang, W. G.; Zhao, Y. P. Profile of volatile compounds in 12 Chinese soy sauces produced by a high-salt-diluted state fermentation. J. Inst. Brew. 2010, 116 (3), 316−328. (3) Lioe, H. N.; Selamat, J.; Yasuda, M. Soy sauce and its umami taste: a link from the past to current situation. J. Food Sci. 2010, 75 (3), 71−76. (4) Lioe, H. N.; Apriyantono, A.; Takara, K.; Wada, K.; Naoki, H.; Yasuda, M. Low molecular weight compounds responsible for savory taste of Indonesian soy sauce. J. Agric. Food Chem. 2004, 52, 5950− 5956. (5) Lioe, H. N.; Takara, K.; Yasuda, M. Evaluation of peptide contribution to the intense umami taste of Japanese soy sauces. J. Food Sci. 2006, 71 (3), 277−283. (6) Imamura, M.; Sato, T.; Hatamoto, O. Effects on thresholds for sweet and umami tastants in soy sauce (in Japanese). Nippon Shokuhin Kagaku Kogaku Kaishi 2009, 56 (7), 384−393. (7) Lioe, H. N.; Apriyantono, A.; Takara, K.; Wada, K.; Yasuda, M. Umami taste enhancement of MSG/NaCl mixtures by subthreshold Lα-aromatic amino acids. J. Food Sci. 2005, 70, 401−405. (8) Oka, S.; Nagata, K. Isolation and characterization of neutral peptides in soy sauce. Agric. Biol. Chem. 1974, 38, 1185−1194. (9) Kaneko, S.; Kumazawa, K.; Nishimura, O. Isolation and identification of the umami enhancing compounds in Japanese soy sauce. Biosci., Biotechnol., Biochem. 2011, 75 (7), 1275−1282. (10) Lioe, H. N.; Wada, K.; Aoki, T.; Yasuda, M. Chemical and sensory characteristics of low molecular weight fractions obtained from three types of Japanese soy sauce (shoyu) − koikuchi, tamari and shiro shoyu. Food Chem. 2007, 100, 1669−1677. (11) Keast, R. S. J.; Breslin, P. A. S. An overview of binary taste-taste interactions. Food Qual. Pref. 2002, 14, 111−124. (12) Wishart, S. D. Metabolomics: applications to food science and nutrition research. Trends Food Sci. Technol. 2008, 19, 482−493. (13) Cevallos-Cevallos, J. M.; Reyes-De-Corcuera, J. J.; Etxeberria, E.; Danyluk, M. D.; Rodrick, G. E. Metabolomic analysis in food science: a review. Trends Food Sci. Technol. 2009, 20, 557−566. (14) Pongsuwan, W.; Fukusaki, E.; Bamba, T.; Yonetani, T.; Yamahara, T.; Kobayashi, A. Prediction of Japanese green tea ranking 7322
dx.doi.org/10.1021/jf501173r | J. Agric. Food Chem. 2014, 62, 7317−7322
