Metabolic Visualization Reveals the Distinct Distribution of Sugars and Amino Acids in Rice Koji (2024)

Keywords: rice koji, mass spectrometry imaging, sugars, amino acids

Samples and reagents

Yamadanishiki-type (polishing rate 70%) steamed rice and rice koji samples were acquired from the mainline production in Hakutsuru Sake Brewing Co., Ltd. The rice koji were collected 42 h after fermentation. Standards (sugars, sugar alcohols and amino acids) were purchased from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan), Merck & Co., Inc. (Rockville, MD, USA), Nacalai Tesque, Inc. (Kyoto, Japan), or Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). For the preparation of solvents, methanol (MeOH) was purchased from Kanto Chemical Co., Inc. (Tokyo, Japan), while ultrapure water was obtained through Barnstead GenPure xCAD Plus water purification system by Thermo Fisher Scientific (Tokyo, Japan). The matrix N-(1-naphthyl)ethylenediamine dihydrochloride (NEDC), 2,5-dihydroxybenzoic acid (DHB), and derivatization agent for amino acid imaging 2,4-diphenyl-pyranylium tetrafluoroborate (DPP-TFB) were purchased from Merck. Triethylamine (TEA) for part of the DPP-TFB buffer was obtained from Wako.

Sample preparation for mass spectrometry imaging

Sample sectioning was conducted based on the previous report.¹³⁾ Serial sections of each sample were obtained and attached to Indium Tin Oxide (ITO) glass (100 Ω/m² without anti-peeling coating; Matsunami Glass, Osaka, Japan) using conductive double-sided tape (Shielding Non-woven Fabric Tape; 3 M company St. Paul, MN, USA). For sugar imaging, a solution of the NEDC matrix (7 mg/mL in 50% MeOH) was applied manually using an airbrush (PS-270, GSI Creos, Tokyo, Japan). For amino acid imaging, the DPP-TFB stock solution (10 mg/mL in MeOH) was first prepared. The working solution for derivatization contained 8% of the stock solution in a buffer containing MeOH, ultrapure water, and TEA (6 : 9 : 0.01, v/v/v). The samples were sprayed with the DPP-TFB working solution and incubated for 1.5 h. The matrix DHB (50 mg/mL in 50% MeOH) was then subsequently sprayed before analysis.

Mass spectrometry imaging analysis

Mass spectrometry imaging was performed using iMScope TRIO (Shimadzu, Kyoto, Japan). For sugar analysis, the instrument was operated in negative mode, and spectra were acquired in the range of m/z 115 to 550. One section of each sample was used for the scan analysis. For amino acid imaging, the instrument was operated in positive mode, and the fragment ion m/z 232.11 were targeted for each amino acid precursor. Two serial sections of each sample were used for the targeted MS/MS analysis. Each section was used to analyze nine targeted amino acids. The area of laser shot was shifted 10 μm to either right or downwards of the original area after each analysis. Laser diameter and intensity was set to 2 (arbitrary unit; approximately 25 μm) and 70 (arbitrary unit), respectively. All MSI images were acquired with a 50 μm pitch. Acquired data were then processed using Imaging MS Solution (Shimadzu, Kyoto, Japan). For the comparison of distribution, absolute intensity from the sample with higher value of each metabolites were used as the high limit value for both samples.

Visualization of sugars by mass spectrometry imaging

Rice koji-making induces the production of amylolytic enzymes, resulting in the conversion of starch into simple sugars. Monosaccharides such as glucose, fructose, galactose as well as disaccharides such as sucrose and maltose are commonly found in rice koji.^4,5) The increase of sugar alcohols such as erythritol, arabitol, and xylitol after rice koji-making were also previously reported.^5,6) In this study, different types of sugars, such as monosaccharide and disaccharide, as well as sugar alcohols, were targeted for MSI analysis. However, isomers were not able to be separated using this method, causing some sugars that share the same m/z value could not be distinguished.

In sugar imaging, the NEDC matrix was utilized as it was reported to be able to visualize glucose due to the formation of [M+Cl]⁻ adducted ion.¹⁴⁾ This principle was known to apply to other sugars.¹⁵⁾ Hence, other sugars are theoretically able to be detected with the same method. Detection was confirmed by analyzing sugar and sugar alcohol standards mixed with the NEDC matrix (Supplementary Fig. S1).

The method was then applied to imaging by spraying the NEDC matrix on the sample. Sugar and sugar alcohols were all detected as [M+Cl]⁻ ion (Fig. 1). There are 5 peaks detected; m/z 157 (erythritol), m/z 187 (arabitol, xylitol), m/z 215 (fructose, galactose, glucose, inositol), m/z 217 (mannitol, sorbitol), and m/z 377 (maltose, sucrose). The peak detected at m/z 215 are thought to be mainly originating from glucose, as the major sugar produced by the amylolytic enzymes originating from rice koji is reported to be glucose.¹⁶⁾ However, the possibility of glucose isomers also included in the distribution profile of m/z 215 should be acknowledged.

Based on the relative intensity, all sugars and sugar alcohols were detected higher in rice koji. Most of the peaks showed even distribution on rice koji, apart from m/z 157 and m/z 187, which shows accumulation near the edge of rice koji. These peaks are derived from the sugar alcohols erythritol as well as arabitol and/or xylitol. These sugar alcohols may be derived from hemicellulose abundant in rice grains, especially in the bran layer. However, the rice used in this study has 70% polishing rate, and is believed to have been removed completely of the bran layer.¹⁷⁾ On the other hand, it has been reported that filamentous fungi such as A. oryzae tend to accumulate polyols or sugar alcohols at a low water activity (a_w) environment.¹⁸⁾ Rice koji making is a type of solid-state fermentation (SSF), and as such, has a low a_w environment. The fungi prevent water loss by accumulating compatible solutes such as polyols. Previous studies have shown the tendency of Yamadanishiki rice to accumulate water in the core of rice grain instead of the surface.^19,20) The high accumulation of these polyols on the surface of rice koji may be due to low a_w in that area compared to the inside of rice. Another theory that could be generated is that the fungus produced polyols to function as a “barrier” to prevent the moisture inside from drying out. In this case, the profile of mycelial growth in rice koji may provide additional perspective in future studies.

On the other hand, not all sugar alcohols seem to accumulate on the edges, such as inositol, mannitol, and sorbitol. Inositol distribution could not be differentiated due to being an isomer with other monosaccharides such as glucose, which has m/z 215. In addition, mannitol and sorbitol at m/z 217 appeared to have the same distribution pattern as glucose at m/z 215. There is a possibility that the m/z 217 peak is mainly detected as an isotope of m/z 215, as this peak also appeared on standard detection of glucose and other sugars sharing the same m/z value.

Visualization of amino acids by mass spectrometry imaging

During the rice koji-making process, the koji mold produces proteolytic enzymes to obtain nitrogen necessary for growth. The change of amino acid content after koji fermentation has been reported by several studies.^4,21) In this study, the distribution of amino acids inside rice koji was investigated (Supplementary Table S2). However, several amino acids with isomers such as leucine and isoleucine were not able to be separated using this method.

In amino acid imaging, derivatization agent DPP-TFB was applied before matrix application. On-tissue detection of amino metabolites has been known to be difficult due to the low-ionization efficiency and spectral interferences from the MALDI matrix. Thus, the use of on-tissue derivatization is one of the methods to increase amino metabolites detectability.²²⁾ Derivatized amino acids were detected through MS/MS analysis as m/z 232.11 after fragmentation of the parent ions (Fig. 2). There are 14 peaks detected; m/z 304 (alanine), m/z 318 (GABA), m/z 320 (serine), m/z 332 (valine), m/z 334 (threonine), m/z 336 (cysteine), m/z 346 (leucine, isoleucine), m/z 347 (asparagine), m/z 348 (aspartic acid), m/z 361 (lysine, glutamine), m/z 362 (glutamic acid), m/z 370 (histidine), m/z 380 (phenylalanine), and m/z 389 (arginine). Other amino acids, such as glycine, proline, tyrosine, and tryptophan, were also investigated. While glycine was able to be detected, the signals were accumulated on blank spots of the samples (data not shown), leading us to believe that it was not the real distribution of glycine. The remaining amino acids, on the other hand, had a very low signal intensity and thus were hard to visualize.

In general, all amino acids were shown to have higher intensity in rice koji relative to steamed rice except for cysteine (m/z 336) and histidine (m/z 370). Furthermore, the distribution profile on rice koji differed into two patterns. The first pattern exhibited accumulation around the edges (alanine, GABA, serine, valine, leucine/isoleucine, asparagine, aspartic acid, lysine/glutamine, glutamic acid), while the other amino acids (threonine, cysteine, histidine, phenylalanine, arginine) were distributed evenly on the cross-section of rice koji. The accumulation of amino acids near the surface of rice koji could be related to how most proteins in rice are near the surface. A previous study mentioned that rice with a low polishing rate contains more protein, and a higher polishing rate will result in lower protein content.²³⁾ This implies that the proteins are mostly accumulated away from the center of the rice. The rice used in this study had a low 70% polishing rate, which means most of the proteins would be concentrated on the edges of rice. Another study using 70% polishing rate Yamadanishiki rice also showed how the storage proteins are not accumulated in the center of the grain,²⁴⁾ backing up this hypothesis.

The remaining amino acids (threonine, cysteine, histidine, phenylalanine, and arginine) were shown to have an even distribution on the cross-section of rice koji. Cysteine was decreased after koji fermentation, implying that A. oryzae mainly utilized the cysteine that was originally already in rice. A previous study also mentioned that the production of cysteine during rice koji making is lower than other amino acids.²¹⁾ The different distribution with other amino acids may suggest that these amino acids have a different role to A. oryzae growth in SSF. However, current results were not able to provide enough information to explain this phenomenon.

On the other hand, the apparent difference in intensity trend between the previous study and visualization of histidine was observed. It has been reported that rice already contained histidine,²⁵⁾ while the production of histidine by A. oryzae was not as high as other amino acids.²¹⁾ There is a possibility that the steamed rice sample used in this study already had a high amount of histidine. In contrast, the rice koji sample used has originally low histidine from the rice as well as low production from the fungi itself.

At a more detailed observation, the edges of rice koji observable from the optical image showed some white parts that were not related to the cracks in the section. These parts corresponded with the accumulation of amino acids as well as several sugar alcohols on the edges of rice koji (Fig. 3). The difference in color may signify a different form of steamed rice cell that may be changed due to A. oryzae growth. It was also reported that a collapsed structure inside rice koji indicates that the fungi have penetrated and digested the rice for nutrition.²⁶⁾ The white-colored parts may as well be the collapsed structure mentioned. However, visualization of mycelial growth may be needed to prove this theory. Nevertheless, the fact that several metabolites distribution were corresponding with the change in rice structure after rice koji making raises the possibility of A. oryzae growth profile being important to metabolites accumulation.

The distribution profile of sugars, sugar alcohols, and amino acids in rice koji was investigated through mass spectrometry imaging analysis. Previous studies have explored the change in the metabolic profile during rice koji making. This study confirmed the findings as well as adding another perspective by visualizing several changed metabolites distribution in rice koji. MSI analysis revealed unique distribution depending on the metabolites observed, a new finding that could not be obtained using conventional methods. Furthermore, the distribution profile seemed to be related to the different structure of rice koji. The new information provided by the imaging results opened possibilities in understanding not only the process happening during rice koji making process, but also the growth characteristic of Aspergillus fungi in general SSF processes.

Supplementary Data

• 1).J. Fujita, Y. I. Yamane, H. Fukuda, Y. Kizaki, S. Wakabayashi, S. Shigeta, O. Suzuki, K. Ono. Production and properties of phytase and acid phosphatase from a sake koji mold, Aspergillus oryzae.J. Biosci. Bioeng.95: 348–353, 2003. [DOI] [PubMed] [Google Scholar]
• 2).K. Ito, A. Kimizuka, N. Okazaki, S. Kobayashi. Mycelial distribution in rice koji.J. Ferment. Bioeng.68: 7–13, 1989. [Google Scholar]
• 3).A. J. Kim, J. N. Choi, J. Kim, S. B. Park, S. H. Yeo, J. H. Choi, C. H. Lee. GC-MS based metabolite profiling of rice koji fermentation by various fungi.Biosci. Biotechnol. Biochem.74: 2267–2272, 2010. [DOI] [PubMed] [Google Scholar]
• 4).S. Lee, D. E. Lee, D. Singh, C. H. Lee. Metabolomics reveal optimal grain preprocessing (milling) toward rice koji fermentation.J. Agric. Food Chem.66: 2694–2703, 2018. [DOI] [PubMed] [Google Scholar]
• 5).A. J. Kim, J. N. Choi, J. Kim, H. Y. Kim, S. B. Park, S. H. Yeo, J. H. Choi, K. H. Liu, C. H. Lee. Metabolite profiling and bioactivity of rice koji fermented by Aspergillus strains.J. Microbiol. Biotechnol.22: 100–106, 2012. [DOI] [PubMed] [Google Scholar]
• 6).D. Lee, S. Lee, E. Jang, H. Shin, B. Moon, C. Lee. Metabolomic profiles of Aspergillus oryzae and Bacillus amyloliquefaciens during rice koji fermentation.Molecules21: 773–787, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 7).E. Handberg, K. Chingin, N. Wang, X. Dai, H. Chen. Mass spectrometry imaging for visualizing organic analytes in food.Mass Spectrom. Rev.34: 641–658, 2015. [DOI] [PubMed] [Google Scholar]
• 8).S. Taira, K. Uematsu, D. Kaneko, H. Katano. Mass spectrometry imaging: Applications to food science.Anal. Sci.30: 197–203, 2014. [DOI] [PubMed] [Google Scholar]
• 9).K. Gorzolka, J. Kölling, T. W. Nattkemper, K. Niehaus. Spatio-temporal metabolite profiling of the barley germination process by MALDI MS imaging.PLoS ONE11: e0150208, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 10).H. Enomoto, K. Sato, K. Miyamoto, A. Ohtsuka, H. Yamane. Distribution analysis of anthocyanins, sugars, and organic acids in strawberry fruits using matrix-assisted laser desorption/ionization-imaging mass spectrometry.J. Agric. Food Chem.66: 4958–4965, 2018. [DOI] [PubMed] [Google Scholar]
• 11).K. Miyoshi, Y. Enomoto, E. Fukusaki, S. Shimma. Visualization of asparaptine in asparagus (Asparagus officinalis) using MALDI-IMS.Anal. Sci.34: 997–1001, 2018. [DOI] [PubMed] [Google Scholar]
• 12).S. Shimma, T. Sagawa. Microscopy and mass spectrometry imaging reveals the distributions of curcumin species in dried turmeric root.J. Agric. Food Chem.67: 9652–9657, 2019. [DOI] [PubMed] [Google Scholar]
• 13).A. P. Wisman, Y. Tamada, S. Hirohata, K. Gomi, E. Fukusaki, S. Shimma. Mapping haze-komi on rice koji grains using β-glucuronidase expressing Aspergillus oryzae and mass spectrometry imaging.J. Biosci. Bioeng.129: 296–301, 2020. [DOI] [PubMed] [Google Scholar]
• 14).J. Wang, S. Qiu, S. Chen, C. Xiong, H. Liu, J. Wang, N. Zhang, J. Hou, Q. He, Z. Nie. MALDI-TOF MS imaging of metabolites with a N-(1-naphthyl) ethylenediamine dihydrochloride matrix and its application to colorectal cancer liver metastasis.Anal. Chem.87: 422–430, 2015. [DOI] [PubMed] [Google Scholar]
• 15).L. Boutegrabet, B. Kanawati, I. Gebefügi, D. Peyron, P. Cayot, R. D. Gougeon, P. Schmitt-Kopplin. Attachment of chloride anion to sugars: Mechanistic investigation and discovery of a new dopant for efficient sugar ionization/detection in mass spectrometers.Chem. Eur. J.18: 13059–13067, 2012. [DOI] [PubMed] [Google Scholar]
• 16).S. Baba, I. Oguri, M. Fukuzawa, T. Iida, I. Kobayashi, K. Imai. Oligosaccharide production by koji-kin enzymes.J. Brew. Soc. Japan69: 781–783, 1974. [Google Scholar]
• 17).K. Yoshizawa, T. Ishikawa. The properties of sake rice and its treatment.J. Brew. Soc. Japan69: 645–650, 1974. [Google Scholar]
• 18).G. J. G. Ruijter, J. Visser, A. Rinzema. Polyol accumulation by Aspergillus oryzae at low water activity in solid-state fermentation.Microbiology150: 1095–1101, 2004. [DOI] [PubMed] [Google Scholar]
• 19).A. K. Horigane, H. Takahashi, S. Maruyama, K. Ohtsubo, M. Yoshida. Water penetration into rice grains during soaking observed by gradient echo magnetic resonance imaging.J. Cereal Sci.44: 307–316, 2006. [Google Scholar]
• 20).M. Yoshida. Observation of moisture distribution affecting texture of food by MRI.Nippon Shokuhin Kagaku Kogaku Kaishi59: 478–483, 2012. [Google Scholar]
• 21).T. Ito, H. Takahashi, T. Shiga, T. Sato, N. Nakazawa, K. Iwano. Analysis of the amino acid content of sake koji and estimation of the amino acid amount formed by Aspergillus oryzae in sake koji making.J. Brew. Soc. Japan108: 453–460, 2013. [Google Scholar]
• 22).C. Esteve, E. A. Tolner, R. Shyti, A. M. J. M. van den Maagdenberg, L. A. McDonnell. Mass spectrometry imaging of amino neurotransmitters: A comparison of derivatization methods and application in mouse brain tissue.Metabolomics12: 30, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 23).Y. Iemura, Y. Ohta, S. Hara. Effects of rice varieties and polishing ratio on the dissolution of protein from steamed rice.J. Brew. Soc. Japan90: 947–952, 1995. [Google Scholar]
• 24).S. Furukawa, T. Mizuma, Y. Kiyokawa, T. Masumura, K. Tanaka, Y. Wakai. Distribution of storage proteins in low-glutelin rice seed determined using a fluorescent antibody.J. Biosci. Bioeng.96: 467–473, 2003. [DOI] [PubMed] [Google Scholar]
• 25).J. S. Kamara, S. Konishi, T. Sasanuma, T. Abe. Variation in free amino acid profile among some rice (Oryza sativa L.) cultivars.Breed. Sci.60: 46–54, 2010. [Google Scholar]
• 26).M. Yoshii, I. Aramaki. Microstructure of sake-koji.J. Brew. Soc. Japan96: 806–813, 2001. [Google Scholar]

Supplementary Materials