Starch Biosynthesis-Related Genotype of Rice (Wx and Alk): Indicator of Suitability for Sake and Shochu Brewing Correlated with Water Absorption Ratio

Abstract

The influence of the genetic background of rice on its suitability for brewing shochu and sake (Japanese alcoholic beverages) has not been clarified. This study investigates the effects of starch biosynthesis-related genes (Wx and Alk) on the water absorption capacity of commercially available rice samples and thus examines the possibility of using rice genotype as a predictor of suitability for shochu and sake brewing. The 24 examined rice samples were classified into four genotypes: Alk/Wx^a, alk/Wx^a, alk/Wx^b, and alk/wx. No significant differences were observed among these genotypes in terms of nonstarch properties such as grain morphology, as well as the water absorption ratio during the initial absorption phase. However, after 1 h of absorption, significant differences in terms of this ratio were observed (Alk/Wx^a < alk/Wx^a < alk/Wx^b < alk/wx). Starch structural analysis revealed that amylose content increased in the order of alk/wx < alk/Wx^b < alk/Wx^a ≈ Alk/Wx^a and the proportion of short amylopectin chains increased in the order of Alk/Wx^a < alk/Wx^a ≈ alk/Wx^b. Correlation analysis revealed a significant negative correlation between the water absorption ratio at 20-120 min and amylose content. The Wx genotype, through its influence on amylose content, was found to play a major role in determining short-term and steamed-rice water absorption ratios regardless of rice type, thus being a promising descriptor of rice suitability for shochu and sake brewing. Thus, our work provides fundamental insights for cultivar breeding and developing brewing methods tailored to specific rice varieties.

Abbreviations

AAC, apparent amylose content; DNA, deoxyribonucleic acid; DP, degree of polymerization; DSC, differential scanning calorimetry; GBSS, granule-bound starch synthase; HPAEC-PAD, high-performance anion-exchange chromatography in combination with pulsed amperometric detection; HPLC, high-performance liquid chromatography; L/W, length-to-width; PCR, polymerase chain reaction; RVA, rapid visco analysis; SD, standard deviation; SNP, single-nucleotide polymorphism; SSIIa, starch synthase IIa.

INTRODUCTION

The growing global popularity of shochu and sake―traditional Japanese alcoholic beverages―has resulted in their increased export from Japan and production in other countries [1]. These beverages are brewed from steamed rice and rice koji (steamed rice inoculated with koji mold, which is subsequently allowed to grow). During brewing, koji enzymes convert starch into sugars, while yeast converts sugars into ethanol [2], and the fermented mash is pressed to produce sake or distilled to produce shochu. The quality (e.g., taste and aroma) of these beverages depends on the characteristics of raw rice (protein content, mineral content, and starch properties) and balance between saccharification and fermentation during brewing. In particular, the proper absorption of water by raw polished rice is a key factor influencing koji quality and subsequent fermentation efficiency [2].

The water absorption ratio of polished rice determines the moisture content of steamed rice, which, in turn, influences koji mold growth in the rice grains, enzyme activity, saccharification, and fermentation [3]. Insufficient water absorption leaves ungelatinized starch granules, resulting in incomplete steaming [4] and thus suppressing koji mold proliferation and reducing enzyme activity, digestibility, and overall material utilization efficiency. Conversely, excessive water absorption leads to the overdigestion of steamed rice and poor koji mold penetration, disrupting the enzyme balance. Inadequate water absorption can lead to poorly developed koji and thus result in insufficient aroma and taste. Additionally, poor water absorption affects rice digestion, typically resulting in a lighter taste, while excessive water absorption promotes rice dissolution, resulting in a richer taste. As the water absorption ratio of steamed rice influences fermentation stages and final product quality, it must be carefully controlled during rice preparation [3, 5].

The water absorption ratio of rice depends on its physicochemical properties, e.g., protein and moisture contents, grain shape, white core (shinpaku) structure, and starch characteristics (which vary with cultivar and production region), as well as soaking temperature [3, 6, 7, 8, 9]. However, the genetic factors underlying these variations have not been fully clarified.

In Japan, Japonica rice is primarily used to brew sake and shochu, whereas Indica rice imported from Thailand is used to brew awamori, a type of shochu. Most previous studies on water absorption have focused on Japonica rice used for sake brewing [3, 10, 11, 12, 13], whereas research on Indica rice used to produce some shochu varieties is scarce [14, 15]. Iemura et al. [10] reported a correlation between amylose content and maximum water absorption ratio, suggesting that starch properties play an important role in determining water absorption behavior. However, their study mainly examined Japonica rice, leaving the water absorption characteristics of Indica rice, which differs in starch composition and structure, unexplored.

The starch properties (e.g., amylose and amylopectin contents and structures) of Japonica and Indica rices are markedly different, and several variations in starch biosynthesis genes between these cultivars have been identified. Generally, Indica rice has a higher content of amylose (a linear polymer featuring glucose residues linked by α-1,4-glycosidic bonds) than Japonica rice [16]. The differences in amylose content are mainly attributed to the activity of granule-bound starch synthase (GBSS) encoded by the Wx gene [17]. Indica rice typically carries the highly expressed Wx^a allele, Japonica rice the low expressed Wx^b allele, and glutinous rice the nonexpressed wx allele, all of which greatly affect amylose content [18, 19]. Amylose content is also influenced by environmental factors during grain filling, as exemplified by its negative correlation with temperature [20, 21, 22]. Thus, amylose content is determined by both genotype and growing conditions.

Amylopectin (a highly branched polymer composed of α-1,4-linked glucose chains with α-1,6 branch points) is synthesized by branching enzymes and starch synthases, among which starch synthase IIa (SSIIa), encoded by the Alk gene, determines the chain-length distribution of amylopectin. Rice with highly active SSIIa (Alk genotype) contains amylopectin with a higher proportion of long chains (degree of polymerization (DP) = 13-24, L-type), whereas that with weakly active SSIIa (alk genotype) contains amylopectin with more abundant short chains (DP = 6-12, S-type) [23]. Japonica rice carries only the S-type (alk) allele, whereas Indica rice carries both Alk and alk alleles [23].

Rice cultivars can be classified into 10 starch composition groups based on the combinations of these genetic variants [24]. Nakamura et al. [8] analyzed the relationships among Alk and Wx genotypes, starch composition, and rice flour properties, revealing that the equilibrium water absorption ratio after 24-h soaking was negatively and positively correlated with apparent amylose content (AAC) and the proportion of short amylopectin chains, respectively.

In brewing, however, rice is soaked for much shorter times before steaming. The effects of genotype and starch composition on the short-time water absorption behavior of raw rice and the water absorption ratio of steamed rice have not been clarified. Understanding how the genetic background of raw rice affects these factors could contribute to the breeding of suitable cultivars and improvement of sake and shochu quality. In particular, the use of genetically diverse rice cultivars, both in and outside Japan, could expand the diversity and global appeal of these traditional beverages. Therefore, one should aim to elucidate the relationship between the water absorption characteristics of rice in the brewing process and the genotype-determined composition and structure of rice starch (e.g., the ratio of amylose to amylopectin and the structure of amylopectin).

Amylose content is traditionally determined using the iodine colorimetric method [25], and starch structure analysis is commonly conducted via treatment with debranching enzymes followed by gel filtration high-performance liquid chromatography (HPLC), high-performance anion-exchange chromatography in combination with pulsed amperometric detection (HPAEC-PAD), or capillary electrophoresis [24]. Other methods for investigating starch properties include gelatinization and pasting property analyses such as differential scanning calorimetry (DSC) and rapid visco analysis (RVA). These methods can be used to probe water absorption capacity. The practical applicability of the corresponding findings can be enhanced through the use of commercially available rice cultivars as samples.

The influence of the genetic background of rice on its suitability for brewing shochu and sake has not been clarified. Herein, we investigated the effects of starch biosyn­thesis-related genes (Wx and Alk) on the water absorption characteristics of commercially available rice samples distributed in and outside Japan and thus examined the possibility of using rice genotype as a predictor of suitability for shochu and sake brewing. Twenty-four rice samples were classified according to their genotypes, and their grain morphologies, starch properties, and water absorption ratios before and after steaming were examined. The relationships among water absorption ratio, genotype, and starch properties were analyzed statistically. Wx genotypes were found to markedly affect the amylose content, short-period water absorption ratio, and water absorption ratio of steamed rice (which are critical factors in brewing) regardless of rice type.

MATERIALS AND METHODS

Materials

The three most common rice (Oryza sativa L.) types are Japonica (round and thick grains), Indica (long and thin grains), and Javanica (big and rather round grains) [26, 27], with each having nonglutinous and glutinous varieties [27]. The 24 employed rice samples, including Indica, Japonica, Javanica, Japonica-Indica hybrid, and Japonica-glutinous varieties, are listed in Table 1. These samples were confirmed to be uncontaminated through deoxyribonucleic acid (DNA) checks and DSC analysis. IR-8, IR-36, and IR-64 cultivars were harvested at the paddy field of the National Agricultural Research Organization, Tsukuba, Japan. Nipponbare and Kasalath cultivars were harvested at the paddy field of the National Research Institute of Brewing, Hiroshima, Japan. The other 19 samples, harvested in Japan, Thailand, India, Pakistan, Spain, and Italy in 2016 and 2019, were purchased from commercial suppliers. Considering environmental factors during cultivation, we also analyzed samples harvested in different years or areas for several varieties. White rice grains with a polishing ratio of ~90 % (polished to 90 % of their original weight) were used. The moisture content of polished rice grains was determined by measuring their weight decrease after heating at 135 °C for 3 h according to a previously reported method [28] and then adjusted to 14.0 wt% in a humidity chamber (KCL-2000, Tokyo Rikakikai Co., Ltd., Tokyo, Japan).

Table 1. Alk/Wx genotypes of rice used in this study.

Sample＊	Type	Cultivation area	Alk genotype	Wx Genotype＊＊
IR-36	Indica	Japan (Ibaraki)	Alk	Wxa
Basmati	Indica	Pakistan	Alk	Wxa (Wxlv)
Kasalath	Indica	Japan (Hiroshima)	Alk	Wxa (Wxlv)
IR-64	Indica	Japan (Ibaraki)	Alk	Wxa (Wxlv)
IR-8	Indica	Japan (Ibaraki)	alk	Wxa
Yumetoiro (2)	Indica	Japan (Chiba, Okayama)	alk	Wxa
Hoshiyutaka (2)	Indica-Japonica hybrid	Japan (Saga)	alk	Wxa (Wxlv)
Vialone	Javanica	Italy	alk	Wxa (Wxlv)
Carnaroli	Javanica	Japan (Ishikawa)	alk	Wxa (Wxlv)
Jasmine (3)	Indica	Thailand	alk	Wxb
Jasmine	Indica	China	alk	Wxb
Nipponbare (2)	Japonica	Japan (Hiroshima)	alk	Wxb
Thai	Indica	Thailand	alk	Wxb
Sari Queen	Indica	Japan (Chiba)	alk	Wxb
Milky Queen	Japonica	Japan (Chiba)	alk	Wxb
Valencia	Javanica	Spain	alk	Wxb
Hiyokumochi (2)	Japonica glutinous	Japan (Hiroshima, Kumamoto)	alk	wx
Himenomochi	Japonica glutinous	Japan (Iwate)	alk	wx

^＊Numbers in parentheses are sample counts.

^＊＊As Wx^lv genotypes belong to the same high-GBSS-activity type as Wx^a ones [19], the former were classified into the same group as the latter in result analysis.

White rice grains were milled into a fine powder using an automated crusher (AC1A, Satake Co., Hiroshima, Japan). The ground material was passed through a 300 μm-mesh sieve and stored at 15 °C until use. Starch was extracted from rice flour using the alkali extraction method [29].

DNA extraction and single-nucleotide polymorphism (SNP) genotyping

Total DNA was extracted from raw rice flour (0.5 g) using the genetically modified organism DNA extraction kit (Nippon Gene Co., Ltd., Tokyo, Japan) according to the manufacturer's protocol and used for the polymerase chain reaction (PCR)-based amplification of target regions in the Alk and Wx genes. Seven primer pairs were designed or adapted from previous reports [30, 31, 32, 33, 34] (Table S1; see J. Appl. Glycosci. Web site) to identify allelic variants associated with the Alk and Wx genotypes (Wx^lv, Wx^a, Wx^b, wx). As Wx^lv genotypes belong to the same high-GBSS-activity type as Wx^a ones [19], the former were classified into the same group as the latter in result analysis. PCR amplifications were performed using the KOD FX Neo kit (TOYOBO Co., Ltd., Osaka, Japan) according to the manufacturer's instructions. SNPs in Alk exon 8 and Wx intron 1 and exon 10 were further genotyped using restriction enzyme digestion [30, 31, 34]. Amplicon sizes were confirmed by agarose gel electrophoresis, and Alk/Wx genotypes were determined.

Grain physicochemical properties

Grain morphology was analyzed using an image-based grain appearance analyzer (Grain Scanner 2 RSQI10B; Satake Co.) after adjusting the grain moisture content to 14.0 wt%. For each sample, ~1,000 kernels were measured, and grain area (mm²), length (mm), width (mm), length-to-width (L/W) ratio, perimeter (mm), and circularity (= 4 × area/(perimeter)²) were recorded. The results were expressed as means ± standard deviations.

Nitrogen content was determined according to the Dumas combustion method following AOAC Official Method 990.03 [35] using an automated elemental analyzer (vario EL III; Elementar Analysensysteme GmbH, Hanau, Germany) and converted to protein content using a factor of 5.95. The above measurements were performed at least twice, and the results were reported as the corresponding means.

Starch structural analysis

AAC was determined by a modified iodine-binding spectrophotometric method [25]. Milled rice flour (100 mg) was gelatinized with ethanol (1 mL) and 1 N NaOH (9 mL), neutralized with 1 N acetic acid (0.5 mL), reacted with an iodine solution (2 % KI + 0.2 % I₂), and diluted to a volume of 50 mL with distilled water. Absorbance was measured at 620 nm (UV-1900; Shimadzu Corporation, Kyoto, Japan) using waxy rice starch and pure amylose (potato, SigmaAldrich, Inc., St. Louis, MO, USA) as standards.

Gel filtration HPLC can be used to evaluate broad chain-length fractions and simultaneously determine amylose and amylopectin contents, providing a practical method of assessing starch structure and its impact on water absorption. The gel filtration HPLC analysis of isoamylase-debranched starch was performed according to a previously reported method [15] with minor modifications. Rice starch (45 mg) was subjected to alkaline treatment, neutralized, and suspended in 30 mM sodium acetate buffer (pH 3.5, 10 mL). Isoamylase (1,180 U/20 μL) was added, and the mixture was incubated at 40 °C for 24 h. After ethanol-induced precipitation and repeated evaporation under reduced pressure, the dried residue was dissolved in 1 N NaOH (0.5 mL), and the solution was diluted with distilled water (1 mL) and filtered through a 0.45-μm membrane filter. An aliquot of the filtrate (100 μL) was injected into an HPLC system (LC-20AD; Shimadzu Corporation) equipped with two sequentially linked Toyopearl HW55S and HW50S columns (7.8 mm × 300 mm, Tosoh Corporation, Tokyo, Japan). The mobile phase (0.02 N NaOH containing 0.5 % NaCl) was supplied at a flow rate of 0.4 mL/min, and the column was maintained at 40 °C. Fractions (87 μL each, collection interval = 13 s) were collected starting at 13 min 14 s after injection, and 100 fractions were collected in total using a fraction collector (SF-2100; Advantec Co., Ltd., Tokyo, Japan). The collection start time of 13 min 14 s was chosen based on the results of preliminary experiments, in which the elution profile was adjusted so that all relevant peaks could be captured within the 100 fractions collected. Each fraction (50 μL) was neutralized with 0.01 N HCl (250 μL), and carbohydrate content was determined by the phenol-sulfuric acid method [36]. The division of these elution regions was based on previous reports [14, 37]. The boundary between the amylose (FI) and amylopectin (FII) fractions was defined at fraction no. 36 (no. 1-36: amylose, no. 37-95: amylopectin). Within the amylopectin fraction (no. 37-95), two peaks were consistently observed across all samples. The trough between these peaks was used to divide the region into long-chain amylopectin (FIIa', no. 37-57) and short- and intermediate-chain amylopectin (FIIb', no. 58-95).

In addition, the FIIb' region was subdivided into intermediate-chain (FIIb'-i, no. 58-78) and short-chain (FIIb'-s, no. 79-95) regions to assess the proportion of short chains.

HPAEC-PAD is suitable for analyzing amylopectin alone and allows the precise determination of amylopectin chain length (DP). Amylopectin chain-length distribution was also analyzed by HPAEC-PAD following a previously reported method [38] with modifications. Analyses were conducted on an ICS-6000 system (Thermo Fisher Scientific Inc., Waltham, MA, USA) using Carbopac PA1 columns (2 mm × 250 mm) and an NaOH/sodium acetate gradient. The proportion of short chains was expressed as the amount of chains with DP 6-9 divided by that of chains with DP 6-24. The above measurements were performed at least twice, and the results were reported as the corresponding means.

Gelatinization, retrogradation, and pasting properties of rice starch

Gelatinization and pasting property analyses based on DSC and RVA provide properties correlated with starch structure and can be used to probe water absorption capacity.

The gelatinization and retrogradation properties of rice starch were examined using DSC (DSC7000X; Hitachi High-Tech Science Corporation, Tokyo, Japan). Purified starch samples (3.5 mg) were sealed with distilled water (12 μL) in pans and heated from 10 to 120 °C at 5 °C/min, with alumina (20 mg) used as a reference. The peak temperature (T_p, °C) and enthalpy change (ΔH, mJ/mg) were obtained automatically. Retrogradation was characterized using the same method for gelatinized samples stored at 4 °C for 14 days and quantified as

Retrogradation degree (%) = 100 % × ΔH₁/ΔH₂,

where ΔH₁ and ΔH₂ are the gelatinization enthalpies of fresh and stored starch, respectively.

Pasting properties were determined using a Rapid Visco Analyzer (RVA-Tec Master; Perten Instruments Ab, Hägersten, Sweden) following a previously reported method [39] with minor modifications. A 9 wt% (dry basis) starch suspension (28 g) was equilibrated at 50 °C, heated to 95 °C at 5 °C/min, held for 5 min, cooled to 50 °C at 5 °C/min, and held for 6 min under constant stirring (160 rpm). The above measurements were performed at least twice, and the results were reported as the corresponding means.

Water absorption test

Water absorption during soaking and after steaming was quantified according to the Annotated Standard Methods of Analysis for Sake Brewing [40] with modifications. Polished rice (W₀ = 10 g) was placed into a water absorption tube with small holes in the bottom, and the tube was placed in a deep stainless steel tray filled with distilled water at 25 °C. After soaking for 10, 20, 30, 60, and 120 min, the tube was removed and drained by centrifugation (KOKUSAN H103N, rotor RF-110; Kokusan Co., Ltd., Tokyo, Japan) at 1,660 G for 3 min. The weight after soaking (W₁) was measured, and the water absorption ratio was calculated as

Water absorption ratio (%) = 100 % × (W₁ − W₀)/W₀.

For water absorption after steaming, polished rice (W₀ = 10 g) was placed into a stainless soup basket with small holes, soaked in distilled water at 15 °C for 15-20 h, drained, and steamed for 50 min in a steam boiler (M-11, Eishin Electric Co., Ltd., Kanagawa, Japan). The steamed rice was cooled to room temperature, sealed in a plastic bag, equilibrated for 3 h at 15 °C, and weighed (W₂). The absorption ratio was calculated as

Water absorption ratio after steaming (%) = 100 % × (W₂ − W₀)/W₀.

The above measurements were performed at least twice, and the results were reported as the corresponding means.

Statistical analysis

Statistical tests were performed using the Data Analysis ToolPak and Real Statistics add-in programs for Microsoft Excel 2019 (Microsoft Corporation, Redmond, WA, USA). Linear regression results were expressed in terms of Pearson's correlation coefficients. A significance level of p < 0.05 was used. Correlation coefficients (r values) were used as correlation quality and predictive evaluation indicators. For significant differences among more than three groups, statistical analysis was performed by one-way analysis of variance and differences between the means were probed using the Tukey-Kramer test when the F-value was significant. A p-value of 0.05 or less was considered significant.

RESULTS AND DISCUSSION

Rice properties and genotypes

Diverse rice types were included, such as Indica (n = 13), Indica-Japonica hybrid (n = 2), Javanica (n = 3), Japonica (n = 3), and Japonica glutinous (n = 3) rices. Grain morphology (area, length, width, L/W ratio, perimeter, circularity), 1,000-kernel weight, protein content, and Alk and Wx genotypes were analyzed (Tables 1, 2 and S2; see J. Appl. Glycosci. Web site). The samples were classified into four genotypes: Alk/Wx^a, alk/Wx^a, alk/Wx^b, and alk/wx. Indica rice was classified into Alk/Wx^a, alk/Wx^a, and alk/Wx^b genotypes, Indica-Japonica hybrid rice into the alk/Wx^a genotype, Javanica rice into the alk/Wx^a genotype, Japonica rice into the alk/Wx^b genotype, and Japonica glutinous rice into the alk/wx genotype. Significant differences in grain morphology, 1,000-kernel weight, or protein content were observed among rice types but not genotypes (Table 2).

Table 2. Properties of rice according to genotype and cultivar.

		Grain morphology						1,000-kernel weight (g)	Protein content (% dry weight)
		Area (mm2)	Width (mm)	Length (mm)	L/W ratio	Perimeter (mm)	Circularity
Genotype
Alk/Wxa	n = 4	10.05 ± 1.44a	2.01 ± 0.22a	6.37 ± 0.99a	3.26 ± 0.82a	14.49 ± 1.82a	0.62 ± 0.10a	16.18 ± 2.42a	6.98 ± 1.24a
alk/Wxa	n = 7	11.89 ± 2.02a	2.59 ± 0.40a	6.01 ± 0.35a	2.37 ± 0.30a	14.40 ± 0.98a	0.68 ± 0.03a	21.31 ± 5.49a	7.27 ± 0.73a
alk/Wxb	n = 10	11.10 ± 1.29a	2.29 ± 0.42a	6.31 ± 1.10a	2.92 ± 0.92a	14.80 ± 1.83a	0.70 ± 0.12a	19.51 ± 3.24a	7.47 ± 0.68a
alk/wx	n = 3	10.22 ± 0.15a	2.74 ± 0.03a	4.94 ± 0.05a	1.81 ± 0.03a	12.54 ± 0.12a	0.82 ± 0.00a	20.13 ± 0.05a	7.23 ± 0.33a
Types
Indica	n = 13	10.88 ± 1.38a	2.08 ± 0.23a	6.63 ± 0.85a	3.26 ± 0.68a	15.11 ± 1.61a	0.63 ± 0.09a	17.89 ± 2.69a	7.36 ± 0.89a
Indica-Japonica hybrid	n = 2	9.94	2.23	5.86	2.64	13.6	0.72	16.12	6.75
Javanica	n = 3	14.25 ± 0.49b	3.13 ± 0.16b	6.02 ± 0.30ab	1.93 ± 0.17b	15.21 ± 0.44ab	0.71 ± 0.05ab	28.61 ± 1.11b	8.13 ± 0.12a
Japonica	n = 3	10.12 ± 0.56a	2.70 ± 0.04b	4.93 ± 0.19b	1.83 ± 0.04b	12.53 ± 0.43b	0.81 ± 0.01b	19.46 ± 0.91b	6.63 ± 0.52a
Japonica glutinous	n = 3	10.22 ± 0.15a	2.74 ± 0.03b	4.94 ± 0.05b	1.81 ± 0.03b	12.54 ± 0.12b	0.82 ± 0.00b	20.13 ± 0.05b	7.23 ± 0.33a

Data are presented as the means ± standard deviations (SDs) of n measurements.

Different letters (a, b) in the same column indicate significant differences among genotypes (p < 0.05).

Japonica and Japonica glutinous rices showed no significant differences in any grain characteristic. In contrast, Indica rice significantly differed from Japonica rice in terms of all properties except grain area. Indica rice grains were the narrowest and longest, as reflected by their largest L/W ratio, which indicated a slender shape. Javanica rice exhibited the largest grain area and highest 1,000-kernel weight while showing the greatest grain width, although not significantly different from that of Japonica and Japonica glutinous rices. Japonica and Japonica glutinous rices showed the smallest L/W ratio, which suggested a shape closest to circular. Protein content did not significantly differ among the rice types.

Water absorption ratio

The water absorption ratios of polished rice obtained from the 24 samples were measured at soaking times of 10-120 min (Table S3; see J. Appl. Glycosci. Web site). Figure 1A shows the mean water absorption ratios obtained at different steeping times for each genotype. In all cases, the water absorption ratio sharply increased within 10-30 min and then gradually increased until 60 min, saturating at 120 min.

Fig. 1. Effects of steeping time on the water absorption ratio of rice grains.

　(A) Changes in water absorption ratio over time (10-120 min) for Alk/Wx^a, alk/Wx^a, alk/Wx^b, and alk/wx. (B-F) Water absorption ratios after steeping for 10, 20, 30, 60, and 120 min, respectively. (G) Water absorption ratio after steaming.

　Alk/Wx^a (n = 4); alk/Wx^a (n = 7); alk/Wx^b (n = 10); alk/wx (n = 3).

　For alk/wx, only two samples were available at 10 and 30 min (n = 2).

　Different letters (a-c) in the same column indicate significant differences among genotypes (p < 0.05).

At 10 min, variations were observed among samples within each genotype but no significant differences were detected between genotypes (Fig. 1B). As the soaking time increased (20-120 min), differences among genotypes emerged and became more pronounced (Figs. 1C-F). At 20 min, alk/wx showed a significantly higher absorption ratio than the other genotypes (Fig. 1C). At 30 min, alk/Wx^b exhibited a significantly higher absorption ratio than Alk/Wx^a (Fig. 1D). At 60 and 120 min, the absorption ratio followed the order of Alk/Wx^a < alk/Wx^a < alk/Wx^b < alk/wx (Figs. 1E and F). No significant difference was found between Alk/Wx^a and alk/Wx^a; however, both groups showed significantly lower absorption ratios than alk/Wx^b, which, in turn, showed significantly lower absorption ratios than alk/wx.

The water absorption ratio of steamed rice measured after overnight soaking was 10.3-16.9 % higher than that measured for polished rice after 120 min (although no significant differences in this increase were observed among genotypes) but followed the same order (Alk/Wx^a < alk/Wx^a < alk/Wx^b < alk/wx) (Fig. 1G, Table S3; see J. Appl. Glycosci. Web site).

These findings suggest that the Wx genotype strongly influenced the water absorption capacity of rice grains during 20-120 min of soaking and after steaming. Furthermore, no significant differences in 60-min, 120-min, and steamed-rice water absorption ratios were found between Indica alk/Wx^b varieties and Japonica alk/Wx^b varieties (Table S3; see J. Appl. Glycosci. Web site).

Correlation analysis (Table 3) revealed that the water absorption ratios at 10 and 20 min were strongly negatively correlated with grain length and L/W ratio (r = −0.79 to −0.59, p < 0.05) and strongly positively correlated with width, circularity, and 1,000-kernel weight (r = 0.56-0.83, p < 0.05). For the maximum water absorption ratio at 120 min, the absolute values of most correlation coefficients significantly decreased. For example, the correlations of grain length and L/W ratio weakened to r = −0.39 and −0.41, respectively, and those of width, circularity, and 1,000-kernel weight weakened to r = 0.42, 0.45, and 0.24, respectively.

Table 3. Coefficients describing correlations between the water absorption ratios and selected properties of rice grains.

	Water absorption ratio (%)
	Steeping period					After steaming
	10 min	20 min	30 min	60 min	120 min
Area (mm2)	0.36	0.27	0.09	−0.12	−0.02	−0.02
Width (mm)	0.81＊	0.83＊	0.64＊	0.47＊	0.42＊	0.34
Length (mm)	−0.59＊	−0.60＊	−0.60＊	−0.56＊	−0.39	−0.30
L/W ratio	−0.79＊	−0.78＊	−0.65＊	−0.52＊	−0.41＊	−0.31
Perimeter (mm)	−0.43＊	−0.42＊	−0.47＊	−0.49＊	−0.30	−0.22
Circularity	0.77＊	0.78＊	0.68＊	0.57＊	0.45＊	0.35
1,000-kernel weight (g)	0.56＊	0.57＊	0.40	0.21	0.24	0.21
Protein content (%/dry)	−0.31	−0.21	−0.17	−0.05	0.03	0.07

^＊p < 0.05 (n = 24).

Previous studies on sake brewing have reported that the internal structures of rice grains, such as shinpaku, promote fast water absorption [11]. The characteristics that showed correlations with the water absorption ratio herein may similarly contribute to faster water uptake.

The above results indicate that morphological traits primarily influenced the initial stages of water absorption, whereas their effects became minimal after 20 min or steaming. Instead, the compositional properties of starch, particularly those determined by the Wx genotype, probably played a dominant role in determining water absorption at longer soaking times and after steaming.

Starch structure

Fifteen rice samples representing four genotypic groups were selected for starch structure analysis. As the alk/wx genotype exhibits an extremely low AAC [41], only one sample (n = 1) was selected for this group. For the other genotypes (Alk/Wx^a, alk/Wx^a, and Alk/Wx^b), four, five, and five samples, respectively, were analyzed.

Purified starch was debranched with isoamylase, and the amylose content and amylopectin chain-length distribution were analyzed using gel filtration HPLC. The elution profile was divided into three regions: amylose (FI, no. 1-36), long-chain amylopectin (FIIa, no. 37-57), and intermediate-to-short-chain amylopectin (FIIb, no. 58-95) (Fig. 2A).

Fig. 2. Analysis of isoamylase-debranched starch by gel filtration chromatography.

　(A) Average chain-length distributions of isoamylase-debranched starch for different genotypes. Difference profiles relative to that of Nipponbare calculated by setting the cumulative distribution for fraction no. (B-E) 1-95 and (F-I) 37-95 as 100 %. (B, F) Alk/Wx^a (n = 4); (C, G) alk/Wx^a (n = 5); (D, H) alk/Wx^b (n = 5); (E, I) alk/wx (n = 1).

Table 4 lists the proportions (relative contents) of all peak fractions, revealing large genotypic differences in the proportion of the apparent amylose fraction including super long chains (FI content = 0.2-24.8 %) [37]. Comparison among genotypes showed that FI content was lowest for alk/wx and highest for Alk/Wx^a (Fig. 2A, Table 4). Although FI content was highly correlated with AAC measured by the iodine colorimetric method, the FI contents obtained by gel filtration HPLC were slightly lower but consistent with those reported previously, indicating the accurate quantification of starch composition (Table 4).

Table 4. Structural characteristics of purified starch from rice grains.

Genotype	Sample	Isoamylase debranched gel filtration chromatography										HPAEC-PAD DP6-9/DP6-24 (%)	AAC (%)
		Starch					Amylopectin
		FI (%)			FIIa (%)	FIIb (%)	FIIa' (%)	FIIb' (%)		FIIb'/Flla'	FIIb'-s/FIIb'-i
		no. 1-36	(no. 1-23)	(no. 24-36)	no. 37-57	no. 58-95	no. 37-57	FIIb'-i no. 58-78	FIIb'-s no. 79-95
Alk/Wxa	IR-36	23.5	19.3	4.2	17.5	59.0	22.8	61.2	16.0	3.38	0.26	8.1	32.8
	Basmati	22.9	18.8	4.1	16.7	60.4	21.7	60.4	17.9	3.61	0.30	9.6	31.9
	Kasalath	24.8	19.3	5.4	16.9	58.4	22.4	60.5	17.1	3.46	0.28	7.4	35.4
	IR-64	22.2	18.1	4.1	18.2	59.6	23.5	60.9	15.6	3.26	0.26	8.5	30.5
alk/Wxa	IR-8	24.2	19.0	5.2	18.4	57.4	24.3	58.2	17.5	3.11	0.30	13.8	32.7
	Yumetoiro 2	24.3	19.1	5.2	18.3	57.5	24.1	57.0	18.9	3.15	0.33	13.6	38.1
	Hoshiyutaka 1	23.7	19.1	4.6	16.7	59.6	21.9	57.4	20.7	3.57	0.36	14.1	36.1
	Vialone	20.0	16.4	3.6	18.5	61.5	23.1	57.8	19.1	3.33	0.33	14.5	28.7
	Carnaroli	19.1	15.8	3.3	19.6	61.3	24.2	58.5	17.3	3.13	0.30	14.4	28.6
alk/Wxb	Jasmine 2	12.8	9.9	2.8	20.4	66.9	23.3	56.9	19.8	3.29	0.35	13.8	16.1
	Nipponbare 2	16.9	14.1	2.8	18.4	64.7	22.1	58.5	19.3	3.52	0.33	13.1	21.8
	Sari Queen	13.5	11.4	2.0	18.4	68.2	21.2	59.8	19.0	3.71	0.32	13.0	19.4
	Milky Queen	6.6	5.0	1.5	21.3	72.1	22.8	58.1	19.1	3.39	0.33	13.6	13.1
	Valencia	15.6	12.6	3.1	19.3	65.1	22.8	57.6	19.6	3.38	0.34	13.8	23.3
alk/wx	Hiyokumochi 1	0.2	0.1	0.1	20.4	79.4	20.5	58.8	20.7	3.89	0.35	14.7	0.0
Alk/Wxa	n = 4＊	23.3 ± 0.9a	18.9 ± 0.5a	4.5 ± 0.6a	17.3 ± 0.6a	59.3 ± 0.7a	22.6 ± 0.6a	60.8 ± 0.3a	16.6 ± 0.9a	3.43 ± 0.13a	0.27 ± 0.02a	8.4 ± 0.8a	32.7 ± 1.8a
alk/Wxa	n = 5＊	22.2 ± 2.2a	17.9 ± 1.5a	4.4 ± 0.8a	18.3 ± 0.9ab	59.5 ± 1.8a	23.5 ± 0.9a	57.8 ± 0.6b	18.7 ± 1.2b	3.26 ± 0.17a	0.32 ± 0.02b	14.1 ± 0.4b	32.9 ± 3.8a
alk/Wxb	n = 5＊	13.1 ± 3.6b	10.6 ± 3.1b	2.4 ± 0.6b	19.5 ± 1.1b	67.4b ± 2.7b	22.5 ± 0.7a	58.2 ± 1.0b	19.3 ± 0.3b	3.46 ± 0.15a	0.33 ± 0.01b	13.7 ± 0.3b	18.7 ± 3.7b
alk/wx	n = 1	0.2	0.1	0.1	20.4	79.4	20.5	58.8	20.7	3.89	0.35	14.7	0.0

^＊Data are presented as the means ± SDs of n measurements.

Different letters (a, b) in the same column indicate significant differences among genotypes (p < 0.05).

Figures 2B-E and S1 (see J. Appl. Glycosci. Web site) show differential chromatograms relative to Nipponbare generated to further clarify structural differences. In the FI region (no. 1-36), Alk/Wx^a and alk/Wx^a showed positive deviations, with Alk/Wx^a exhibiting a higher peak intensity than alk/Wx^a. In contrast, alk/Wx^b and alk/wx displayed negative deviations, with the latter showing the strongest negative peak. FI content followed the order of alk/wx < alk/Wx^b < alk/Wx^a < Alk/Wx^a. Among the three groups with n ≥ 3, statistical analysis revealed no significant difference between Alk/Wx^a and alk/Wx^a, although these genotypes featured significantly higher FI contents than alk/Wx^b, i.e., FI content followed the order of alk/Wx^b < alk/Wx^a ≈ Alk/Wx^a (Fig. 3A, Table 4). This trend was consistent with the AAC results obtained by the iodine-binding method (Table 4).

Fig. 3. Distribution of starch structural parameters among different genotypes.

　(A-D) Comparison of starch structural parameters among the four genotypes (Alk/Wx^a, alk/Wx^a, alk/Wx^b, and alk/wx). Parameters shown are (A) FI content, (B) FIIb'-s (no. 79-95) content, (C) FIIb'/FIIa' ratio, and (D) FIIb'-s/FIIb'-i ratio.

　^＊FIIb'-s and FIIb'-i indicate short-chain (no. 79-95) and intermediate-chain (no. 58-78) fractions within the FIIb' region, respectively.

　Alk/Wx^a (n = 4); alk/Wx^a (n = 5); alk/Wx^b (n = 5); alk/wx (n = 1).

　Different letters (a, b) in the same column indicate significant differences (p < 0.05).

Regarding amylopectin chain-length distribution, Alk/Wx^a differed from the other genotypes, particularly in the FIIb region (intermediate and short chains). The differential chromatograms of amylopectin fractions relative to Nipponbare (Figs. 2F-I and S2; see J. Appl. Glycosci. Web site) revealed that Alk/Wx^a showed positive values in the early portion of FIIb (no. 58-78, corresponding to intermediate chains) and negative values in the later portion (no. 79-95, corresponding to short chains). This pattern was unique to Alk/Wx^a and distinct from those observed for the three alk genotypes.

Based on these differences, the FIIb region was subdivided into intermediate-chain (FIIb-i, no. 58-78) and short-chain (FIIb-s, no. 79-95) regions, and their proportions relative to total amylopectin and the FIIb'-s/FIIb'-i ratio were calculated (Table 4). Although the mean FIIb'/FIIa' ratio (ratio of intermediate or short chains to long chains) showed no significant differences among genotypes (Fig. 3C), the content of short chains (FIIb'-s, no. 79-95) and the FIIb'-s/FIIb'-i ratio (Figs. 3B and D) increased in the order of Alk/Wx^a < alk/Wx^a < alk/Wx^b. Although alk/Wx^b and alk/Wx^a did not differ significantly, both had significantly higher FIIb'-s contents and FIIb'-s/FIIb'-i ratios than Alk/Wx^a.

HPAEC-PAD analysis (Fig. S3; see J. Appl. Glycosci. Web site) revealed that the Alk genotype featured a lower content of short amylopectin chains (DP 6-9/DP 6-24) than the alk genotype, as reported previously [42]. This result was similar to that of gel filtration HPLC analysis, indicating that Alk gene expression reduces short-chain content and promotes chain elongation (conversion of short chains to intermediate ones), consistent with previous reports [42, 43]. Based on the length of the branches where extended by Alk (SSIIa) [42], we also calculated the DP 6-11/DP 6-24 ratio. The trends in this ratio generally agreed with those observed for the DP 6-9/DP 6-24 ratio (Fig. S3; see J. Appl. Glycosci. Web site).

Amylose content and the proportion of amylopectin short chains are known to be negatively correlated with temperature during grain filling [20, 21, 22, 44]. Thus, the intragenotypic variations observed in Table 4 probably reflect regional climatic conditions during cultivation, whereas significant differences among genotypes indicate that genetic factors exert a stronger influence than environmental conditions during ripening.

The analysis of correlations between starch structural parameters and the water absorption ratio (Table 5, Fig. 4) revealed time-dependent relationships. At 10 min, the water absorption ratio showed no significant correlations with amylose (FI, r = 0.03), intermediate-chain amylopectin (FIIb'-i, r = −0.10), and short-chain amylopectin (FIIb'-s, r = −0.16) contents. This finding agrees with the data in Table 5 and suggests that early water absorption is primarily affected by nonstarch properties such as grain morphology. Correlations with starch structure parameters emerged with the increasing soaking time. Amylose (FI) content was strongly negatively correlated with the water absorption ratio at 20-120 min and that of steamed rice (Table 5, Figs. 4A and D; r = −0.58 to −0.96). AAC measured by the iodine-binding method showed similar correlations (Table 5, r = −0.57 to −0.95). Such correlations were previously reported for the water absorption ratio at saturation (after 24 h) [8]. However, our results revealed that a significant correlation can also be observed for the 20-min water absorption ratio, which better fits the timescale of typical brewing.

Table 5. Coefficients describing correlations between the water absorption ratios of rice grains and structural characteristics of rice starch.

		Water absorption ratio (%)
		Steeping period					After steaming
		10 min	20 min	30 min	60 min	120 min
Starch
FI (%)	no. 1-36	0.03	−0.58＊	−0.79＊	−0.94＊	−0.96＊	−0.95＊
	no. 1-23	0.04	−0.58＊	−0.79＊	−0.94＊	−0.96＊	−0.95＊
	no. 24-36	0.00	−0.57＊	−0.75＊	−0.89＊	−0.91＊	−0.90＊
FIIa (%)	no. 37-57	0.19	0.74＊	0.86＊	0.85＊	0.85＊	0.84＊
FIIb (%)	no. 58-95	−0.08	0.52＊	0.73＊	0.91＊	0.93＊	0.92＊
	no. 58-78	−0.05	0.50	0.71＊	0.87＊	0.88＊	0.89＊
	no. 79-95	−0.11	0.47	0.67＊	0.86＊	0.89＊	0.86＊
Amylopectin
FIIa' (%)	no. 37-57	0.34	0.12	−0.10	−0.37	−0.37	−0.36
FIIb' (%)
FIIb'-i	no. 58-78	−0.10	−0.33	−0.32	−0.29	−0.35	−0.30
FIIb'-s	no. 79-95	−0.16	0.23	0.36	0.54＊	0.59＊	0.54＊
FIIb'/Flla'		−0.33	−0.10	0.11	0.39	0.39	0.38
FIIb'-s/FIIb'-i		−0.12	0.25	0.36	0.50	0.55＊	0.50
DP 6-9/DP 6-24 (%)		0.32	0.58＊	0.59＊	0.59＊	0.62＊	0.56＊
AAC (%)		0.01	−0.57＊	−0.77＊	−0.92＊	−0.94＊	−0.95＊

^＊p < 0.05 (n = 15).

Fig. 4. Relationships between starch structure and water absorption ratios after 120 min of steeping and steaming.

　(A, D) Relationships between FI content and water absorption ratios after (A) 120 min of steeping and (D) steaming. (B, C, E, F) Relationships between FIIb' content and water absorption ratios after steeping or steaming. FIIb' was divided into (B, E) no. 58-78 and (C, F) no. 79-95.

In contrast, amylopectin chain-length features exhibited weaker correlations: FIIb'-i content showed weak negative correlations (r = −0.33 to −0.30; Table 5, Figs. 4B and E), while FIIb'-s content showed moderate positive correlations (r = 0.23 to 0.59, Table 5, Figs. 4C and F), as did the short-chain ratio (DP 6-9/DP 6-24; r = 0.56-0.59; Table 5). However, these were weaker than the correlations with amylose content. Nakamura et al. [8] suggested that amylopectin short-chain ratio may affect saturated water absorption, although our results indicate that its contribution is minor.

Therefore, even under conditions typical of brewing (soaking time = 20-120 min), amylose content, determined by the Wx genotype, exerted a dominant influence on the water absorption ratios of rice grains and steamed rice.

Almost all rice used as a raw material for Japanese sake brewing is of the Japonica variety with the alk/Wx^b genotype. Although we examined only two Japonica varieties with this genotype, empirical and experimental evidence suggests differences in water absorption ratios among the Japonica varieties used for sake brewing. Regarding the underlying reason, Iemura et al. [10] reported a negative correlation between the maximum water absorption rate and amylose content of Japonica rice varieties used for sake brewing (table and sake-type rice), in line with the present results (Table 5). However, for Japonica sake-type rice varieties, water absorption did not show a significant correlation with amylose content, which suggests the influence of factors such as the endosperm (shinpaku) structure. Therefore, the factors affecting the water absorption characteristics of sake-type rice varieties have not been clarified despite their practical significance and should be determined in future works.

Gelatinization and pasting properties of starch

DSC analysis and RVA were performed on 24 rice starch samples to investigate the relationship between starch thermal and pasting properties and the water absorption ratio (Table 6).

Table 6. Gelatinization and pasting properties of rice starch according to genotype.

Genotype	Sample	DSC			Retrogradation degree (%)	RVA
		Tp (°C)	ΔH1 (mJ/mg)	ΔH2 (mJ/mg)		Peak viscosity (cP)	Trough (cP)	Breakdown (cP)	Final viscosity (cP)	Setback (cP)	Peak time (min)	Pasting temperature (°C)
Alk/Wxa	IR-36	73.9	10.9	7.9	74.0	2,987.5	1,124.5	1,863.0	3,216.0	2,091.5	9.6	74.8
	Basmati	70.0	10.2	6.2	57.6	2,766.0	830.0	1,936.0	2,777.0	1,947.0	9.5	71.6
	Kasalath	72.9	10.8	7.7	72.2	2,158.0	695.5	1,462.5	2,325.5	1,630.0	9.7	74.6
	IR-64	73.4	11.7	7.4	68.8	2,998.5	860.0	2,138.5	2,576.5	1,716.5	9.4	74.6
alk/Wxa	IR-8	62.4	9.4	3.5	37.0	2,738.0	1,621.0	1,117.0	4,727.5	3,106.5	10.4	68.3
	Yumetoiro 1	60.2	8.4	3.0	35.8	2,848.0	1,543.0	1,305.0	4,936.5	3,393.5	10.4	68.3
	Yumetoiro 2	63.8	8.5	2.9	35.8	3,135.5	1,562.5	1,573.0	4,936.0	3,373.5	9.8	67.6
	Hoshiyutaka 1	60.8	9.0	2.0	22.8	1,791.0	691.0	1,100.0	2,479.5	1,788.5	10.1	70.4
	Hoshiyutaka 2	58.8	8.3	1.6	18.7	1,805.5	615.0	1,190.5	2,376.5	1,761.5	9.9	69.6
	Vialone	66.6	10.7	2.7	26.2	2,639.0	1,169.0	1,470.0	3,082.5	1,913.5	10.1	70.6
	Carnaroli	68.3	10.5	3.0	26.4	2,873.5	1,349.5	1,524.0	3,203.5	1,854.0	10.3	72.6
alk/Wxb	Jasmine 1	67.9	11.7	1.5	11.9	3,952.0	919.0	3,033.0	1,729.0	810.0	8.6	67.6
	Jasmine 2	67.6	13.0	1.5	11.3	4,141.0	1,133.0	3,008.0	2,048.5	915.5	8.8	67.6
	Jasmine 3	66.8	11.8	1.9	17.2	3,804.5	1,087.5	2,717.0	2,037.5	950.0	9.6	67.6
	Jasmine 4	66.5	11.7	1.1	9.7	4,026.5	991.0	3,035.5	2,032.0	1,041.0	8.2	65.6
	Nipponbare 1	66.5	11.2	1.3	11.6	3,262.0	1,141.0	2,121.0	2,812.0	1,671.0	9.0	66.8
	Nipponbare 2	66.6	9.2	1.4	14.5	3,120.5	930.0	2,190.5	2,348.0	1,418.0	9.9	67.8
	Thai	68.0	10.4	2.0	18.3	3,552.0	1,037.0	2,515.0	2,269.5	1,232.5	9.8	68.6
	Sari Queen	68.1	10.9	1.7	15.7	4,090.0	1,302.5	2,787.5	2,798.5	1,496.0	9.3	68.6
	Milky Queen	69.2	12.9	0.9	6.8	4,471.0	1,196.5	3,274.5	1,818.5	622.0	7.8	66.6
	Valencia	66.3	10.4	1.6	14.7	3,228.5	1,165.5	2,063.0	2,413.0	1,247.5	9.8	67.9
alk/wx	Hiyokumochi 1	64.5	14.8	0.0	0.0	2,476.5	893.5	1,583.0	1,448.5	555.0	5.7	61.9
	Himenomochi 2	64.5	13.8	0.0	0.0	2,314.5	831.0	1,483.5	1,321.5	490.5	5.9	61.6
Alk/Wxa	n = 4＊	72.5 ± 1.5a	10.9 ± 0.5ab	7.3 ± 0.7a	68.2 ± 6.4a	2,727.5 ± 341.6a	877.5 ± 155.5a	1,850 ± 245.4a	2,723.8 ± 326.1ab	1,846.3 ± 183.0ab	9.5 ± 0.1ab	73.9 ± 1.3a
alk/Wxa	n = 7＊	63.0 ± 3.2b	9.2 ± 0.9a	2.7 ± 0.6b	28.9 ± 6.7b	2,547.2 ± 494.2a	1,221.6 ± 386.9a	1,325.6 ± 183.0a	3,677.4 ± 1067.6a	2,455.9 ± 729.8a	10.1 ± 0.2a	69.6 ± 1.6b
alk/Wxb	n = 10＊	67.3 ± 0.9c	11.3 ± 1.1b	1.5 ± 0.3c	13.2 ± 3.4c	3,764.8 ± 430.3b	1,090.3 ± 115.8a	2,674.5 ± 410.4b	2,230.7 ± 352.2b	1,140.4 ± 313.0b	9.1 ± 0.7b	67.5 ± 0.9c
alk/wx	n = 2	64.5	14.3	0	0	2,395.5	862.3	1,533.3	1,385.0	522.8	5.8	61.7

^＊Data are presented as the means ± SDs of n measurements.

Different letters (a-c) in the same column indicate significant differences among genotypes (p < 0.05). cP, centipoise.

The mean T_p was the highest for Alk/Wx^a (72.5 ± 1.5 °C) and lower for genotypes carrying the alk allele (alk/wx: 64.5 °C, alk/Wx^b: 67.3 ± 3.2 °C, alk/Wx^a: 63.0 ± 3.2 °C) (Table 6). The difference in T_p between Alk and alk genotypes notably exceeded the variation among the Wx types. T_p was not significantly correlated with the water absorption ratio at any time point for raw or steamed rice (r = −0.19 to −0.05, Table 7).

Table 7. Coefficients describing correlations between the water absorption ratios of rice grains and gelatinization and pasting properties of rice starch.

	Water absorption ratio (%)
	Steeping period					After steaming
	10 min	20 min	30 min	60 min	120 min
DSC
Tp (°C)	−0.12	−0.11	−0.05	−0.09	−0.17	−0.09
ΔH1 (mJ/mg)	−0.15	0.38	0.62＊	0.74＊	0.76＊	0.80＊
ΔH2 (mJ/mg)	−0.04	−0.46＊	−0.61＊	−0.69＊	−0.73＊	−0.68＊
Retrogradation degree (%)	−0.02	−0.47＊	−0.64＊	−0.73＊	−0.77＊	−0.73＊
RVA
Peak viscosity (cP)	−0.26	0.04	0.13	0.12	0.21	0.26
Trough (cP)	0.40	0.28	0.09	−0.10	−0.07	−0.12
Breakdown (cP)	−0.43＊	0.07	0.11	0.17	0.25	0.33
Final viscosity (cP)	0.34	−0.12	−0.40	−0.57＊	−0.58＊	−0.62＊
Setback (cP)	0.29	−0.24	−0.51＊	−0.66＊	−0.69＊	−0.73＊
Peak time (min)	0.17	−0.35	−0.61＊	−0.80＊	−0.83＊	−0.86＊
Pasting temperature (°C)	0.01	−0.44＊	−0.63＊	−0.77＊	−0.81＊	−0.80＊

^＊p < 0.05 (n = 23).

In contrast, the gelatinization enthalpy (ΔH₁) markedly differed among Wx genotypes within the alk group, following the order of alk/Wx^a < alk/Wx^b < alk/wx (Table 6). As ΔH₁ represents the energy required to melt the crystalline structure of amylopectin [45], samples with lower FI contents (i.e., lower amylose and higher amylopectin contents) tended to show higher ΔH₁ values. Although no significant difference was observed between Alk/Wx^a and alk/Wx^a, the ΔH₁ values of the Alk genotype tended to be higher, possibly because Alk/Wx^a contained amylopectin with more intermediate chains (FIIb-i, no. 58-78) and fewer short chains (FIIb-s, no. 79-95) (Fig. 2F), requiring greater energy for gelatinization (Tables 4 and 6). Therefore, both Wx and Alk genotypes probably influenced ΔH₁. Furthermore, ΔH₁ showed strong positive correlations with water absorption ratios at 30, 60, and 120 min and the water absorption ratio of steamed rice (r = 0.62-0.80) (Table 7). This pattern was similar to that observed for amylose content (Figs. 4A and D), suggesting that ΔH₁ may serve as an indicator of water absorption capacity at longer soaking times or after steaming.

RVA revealed significant differences among genotypes in terms of peak viscosity, breakdown viscosity, final viscosity, setback viscosity, peak time, and pasting temperature (Table 6). Among these parameters, breakdown viscosity showed a moderate negative correlation with the water absorption ratio at 10 min (r = −0.43) (Table 7). For longer soaking times (≥ 60 min) and steamed rice, significant correlations (|r| = 0.57-0.86) were observed for final viscosity, setback viscosity, peak time, and pasting temperature, indicating that these parameters may also reflect water absorption behavior under brewing conditions (Table 7).

Final viscosity, setback viscosity, and peak time followed the order of alk/wx < alk/Wx^b < Alk/Wx^a ≈ alk/Wx^a, the inverse of that observed for amylose (FI) content. Peak viscosity followed the order of alk/wx < alk/Wx^a ≈ Alk/Wx^a < alk/Wx^b, while breakdown viscosity followed the order of alk/Wx^a < alk/wx < Alk/Wx^a < alk/Wx^b, with significant differences observed between Wx^a and Wx^b (except alk/wx). Pasting temperature increased in the order of alk/wx < alk/Wx^b < alk/Wx^a < Alk/Wx^a, with Alk/Wx^a showing the highest value (73.9 ± 1.3 °C) and alk genotypes showing the lowest values (61.7-69.6 ± 1.6 °C, Table 7). This pattern paralleled the T_p trend and may reflect differences in the proportion of short amylopectin chains.

Overall, the results of DSC analysis and RVA indicated that the thermal and pasting parameters of rice starch are closely correlated with water absorption ratios at soaking times of  > 30 min and after steaming, thus providing potential indicators of water uptake behavior relevant to brewing suitability.

CONCLUSION

This study investigated the effects of rice grain morphology, composition, and starch biosynthesis-related genotypes (Alk and Wx) on water absorption characteristics, which are critical factors determining the brewing property of rice. The 24 examined rice samples were classified into four genotypes (Alk/Wx^a, alk/Wx^a, alk/Wx^b, alk/wx), which showed no signif­icant differences in terms of grain morphology, 1,000-kernel weight, or protein content. In contrast, water absorption ratios at soaking times above 20 min were higher for alk/wx. No significant differences were detected during the initial soaking period, which suggests that early water absorption is primarily affected by nonstarch properties such as grain morphology. Significant differences emerged after 1 h and after steaming, with the water absorption ratio following the order of Alk/Wx^a ≈ alk/Wx^a < alk/Wx^b < alk/wx. Furthermore, no significant difference between Indica alk/Wx^b varieties and Japonica alk/Wx^b varieties was found in terms of the water absorption ratios at 60 min, 120 min, and after steaming.

Regarding starch composition and structure, amylose content increased in the order of alk/wx < alk/Wx^b < alk/Wx^a ≈ Alk/Wx^a, and the proportion of short amylopectin chains increased in the order of Alk/Wx^a < alk/Wx^a ≈ alk/Wx^b. The water absorption ratios observed at 20-120 min after soaking and for steamed rice were significantly correlated with amylose content. Furthermore, the thermal and pasting properties of starch, which are closely related to starch composition and structure, also showed strong correlations with the water absorption ratio, thus holding promise as indicators of water absorption behavior.

Overall, our results demonstrate that despite differences in cultivation regions, the starch properties of commercial rice varieties can be effectively classified according to the Alk and Wx genotypes. Regardless of rice type, the Wx genotype, through its influence on amylose content, plays a major role in determining short-term and steamed-rice water absorption ratios, which are key factors affecting the suitability of rice for sake and shochu brewing.

CONFLICT OF INTERESTS

The authors declare no conflict of interests.

ACKNOWLEDGEMENTS

We thank Dr. Umemoto T. (National Agricultural Research Organization) for providing rice samples (IR-8, IR-36, IR-64) and advice regarding the method of genotype grouping (Alk/Wx).

REFERENCES

• [1]  National Tax Agency Japan [Internet]. 2025 [cited 2025 Jul 10]. Available from: https://www.nta.go.jp/taxes/sake/yushutsu/yushutsu_tokei/index.htm.
• [2]  Okuda M. Rice used for Japanese sake making. Biosci Biotechnol Biochem. 2019; 83: 1428-41.
• [3]  Kumagai C, Kuroyanagi Y, Noshiro K. Studies on the water absorption of polished rice for sake brewing. I. Relationship between moisture content and water absorption ratio of sake rice. J Brew Soc Japan. 1976; 71: 718-22 (in Japanese).
• [4]  Koura J, Tamaki M, Aramaki I, Inatani T. Differences in gelatinization properties of rice starch depending on its location in the grain―Studies on steamed rice for sake brewing (Part 4)―. J Brew Soc Japan. 2010; 105: 659-63.
• [5]  Shanthilal J, Anandharamakrishnan C. Computational and numerical modeling of rice hydration and dehydration: a review. Trends Food Sci Technol. 2013; 31: 100-17.
• [6]  Horigane AK, Takahashi H, Maruyama S, Ohtsubo KI, Yoshida M. Water penetration into rice grains during soaking observed by gradient echo magnetic resonance imaging. J Cereal Sci. 2006; 44: 307-16.
• [7]  Lapčíková B, Lapčík L, Valenta T, Majar P, Ondroušková K. Effect of the rice flour particle size and variety type on water holding capacity and water diffusivity in aqueous dispersions. LWT. 2021; 142: 111082.
• [8]  Nakamura M, Mizukami Y, Aoki N, Umemoto T, Hiwatashi M, Ikeda T, et al. Correlation of type of starch composition with milling properties, water absorption and shape of starch granules of rice. Jpn J Crop Sci. 2014; 83: 126-35 (in Japanese).
• [9]  Tomita H, Fukuoka M, Takemori T, Sakai N. Development of the visualization and quantification method of the rice soaking process by using the digital microscope. J Food Eng. 2019; 243: 33-8.
• [10]  Iemura Y, Kageyama Y, Matsunaga K, Hara S. Effects of amylose content, white-core, and white-belly on the water-absorption capacity of rice grains. Studies on quality evaluation of rice for sake brewing (Part 3). J Brew Soc Japan. 1996; 91: 515-20 (in Japanese).
• [11]  Yanagiuchi T, Yamamoto H, Miyazaki N, Nagano T, Wakai Y. Influence of grain type on physical and chemical properties of rice for sake brewing. Seibutsu Kogaku. 1996; 74: 97-103 (in Japanese).
• [12]  Mizuma T, Tomita A, Kitaoka A, Kiyokawa Y, Wakai Y. Water-absorption ratio equation of rice for brewing sake. J Biosci Bioeng. 2007; 103: 60-5.
• [13]  Mizuma T, Kiyokawa Y, Wakai Y. Water absorption characteristics and structural properties of rice for sake brewing. J Biosci Bioeng. 2008; 106: 258-62.
• [14]  Okuda M, Joyo M, Fukuda H, Tamamura T. Effects of moisture and pulverizing of rice on the absorption rate of rice grains used for awamori; improvement of the absorption ratio of steamed rice by two steaming. J Brew Soc Japan. 2017; 112: 629-40 (in Japanese).
• [15]  Okuda M, Joyo M, Fukuda H, Tamamura T. Characteristics of the starch and protein composition of rice grains used for awamori. J Brew Soc Japan. 2019; 114: 585-95 (in Japanese).
• [16]  Cheng FM, Zhong LJ, Wang F, Zhang GP. Differences in cooking and eating properties between chalky and translucent parts in rice grains. Food Chem. 2005; 90: 39-46.
• [17]  Sano Y. Differential regulation of waxy gene expression in rice endosperm. Theor Appl Genet. 1984; 68: 467-73.
• [18]  Isshiki M, Morino K, Nakajima M, Okagaki RJ, Wessler SR, Izawa T, et al. A naturally occurring functional allele of the rice waxy locus has a GT to TT mutation at the 5′ splice site of the first intron. Plant J. 1998; 15: 133-8.
• [19]  Zhang C, Zhu J, Chen S, Fan X, Li Q, Lu Y, et al. Wx^lv, the ancestral allele of rice Waxy gene. Mol Plant. 2019; 12: 1157-66.
• [20]  Umemoto T, Terashima K, Nakamura Y, Satoh H. Differences in amylopectin structure between two rice varieties in relation to the effects of temperature during grain-filling. Starch/Stärke. 1999; 51: 58-62.
• [21]  Inouchi N, Ando H, Asaoka M, Okuno K, Fuwa H. Effect of environmental temperature on distribution of unit chains of rice amylopectin. Starch/Stärke. 2000; 52: 8-12.
• [22]  Hirano HY, Sano Y. Enhancement of Wx gene expression and the accumulation of amylose in response to cool temperatures during seed development in rice. Plant Cell Physiol. 1998; 39: 807-12.
• [23]  Nakamura Y, Sakurai A, Inaba Y, Kimura K, Iwasawa N, Nagamine T. The fine structure of amylopectin in endosperm from Asian cultivated rice can be largely classified into two classes. Starch/Stärke. 2002; 54: 117-31.
• [24]  Umemoto T. Analysis method of starch property based on chain length distribution of amylopectin. Jpn J Crop Sci. 2009; 78: 107-12 (in Japanese).
• [25]  Perez CM, Juliano BO. Modification of the simplified amylose test for milled rice. Starch/Stärke. 1978; 30: 424-6.
• [26]  Chang T. The origin, evolution, cultivation, dissemination, and diversification of Asian and African rices. Euphytica. 1976; 25: 425-41.
• [27]  Hoshikawa K. The Growing Rice Plant: An Anatomical Monograph. Tokyo: Nobunkyo; 1989. p. 3-7.
• [28]  Ito K, Yoshida K, Okazaki N, Kobayashi S. Effect of processing on the pore size distribution and digestibility of rice grain. Agric Biol Chem. 1988; 52: 3001-7.
• [29]  Yamamoto K, Sawada S, Onogaki T. Effects of quality and quantity of alkaline solution on the properties of the rice starch. J Jpn Soc Starch Sci. 1981; 28: 241-4.
• [30]  Lu Y, Xiao P, Shao Y, Zhang G, Thanyasiriwat T, Bao J. Development of new markers to genotype the functional SNPs of SSIIa, a gene responsible for gelatinization temperature of rice starch. J Cereal Sci. 2010; 52: 438-43.
• [31]  Yamanaka S, Nakamura I, Watanabe KN, Sato YI. Identification of SNPs in the waxy gene among glutinous rice cultivars and their evolutionary significance during the domestication process of rice. Theor Appl Genet. 2004; 108: 1200-4.
• [32]  Wanchana S, Toojinda T, Tragoonrung S, Vanavichit A. Duplicated coding sequence in the waxy allele of tropical glutinous rice (Oryza sativa L.). Plant Sci. 2003; 165: 1193-9.
• [33]  Ii Y, Adachi S. DNA markers detecting non-glutinous rice in glutinous rice processed food. J Jpn Soc Food Sci Technol. 2013; 60: 705-10 (in Japanese).
• [34]  Li K, Wang LY, Ren H, Ge Y. Genotyping USDA rice (Oryza spp.) mini-core collection with functional markers for important agronomic traits. Genet Resour Crop Evol. 2022; 69: 2331-44.
• [35]  AOAC Official Method 990.03. Official Methods of Analysis of AOAC International. 18th ed. Gaithersburg (MD): AOAC International; 2005.
• [36]  DuBois M, Gilles KA, Hamilton JK, Rebers PA, Smith F. Colorimetric method for determination of sugars and related substances. Anal Chem. 1956; 28: 350-6.
• [37]  Horibata T, Nakamoto M, Fuwa H, Inouchi N. Structural and physicochemical characteristics of endosperm starches of rice cultivars recently bred in Japan. J Appl Glycosci. 2004; 51: 303-13.
• [38]  Nakamura Y, Kubo A, Shimamune T, Matsuda T, Harada K, Satoh H. Correlation between activities of starch debranching enzyme and α-polyglucan structure in endosperms of sugary-1 mutants of rice. Plant J. 1997; 12: 143-53.
• [39]  Jane J, Chen Y, Lee LF, McPherson AE, Wong KS, Radosavljevic M, et al. Effects of amylopectin branch chain length and amylose content on the gelatinization and pasting properties of starch. Cereal Chem. 1999; 76: 629-37.
• [40]  National Research Institute of Brewing. Commentary on standard analytical methods of National Research Institute of Brewing, Japan. The Brewing Society of Japan; 2017. p. 160-90 (in Japanese).
• [41]  Wang ZY, Zheng FQ, Shen GZ, Gao JP, Snustad DP, Li MG, et al. The amylose content in rice endosperm is related to the post-transcriptional regulation of the waxy gene. Plant J. 1995; 7: 613-22.
• [42]  Umemoto T, Yano M, Satoh H, Shoumura A, Nakamura Y. Mapping of a gene responsible for the difference in amylopectin structure between japonica-type and indica-type rice varieties. Theor Appl Genet. 2002; 104: 1-8.
• [43]  Butardo VM Jr, Luo J, Li Z, Gidley MJ, Bird AR, Tetlow IJ, et al. Functional genomic validation of the roles of soluble starch synthase IIa in Japonica rice endosperm. Front Genet. 2020; 11: 289.
• [44]  Asaoka M, Okuno K, Fuwa H. Effect of environmental temperature at the milky stage on amylose content and fine structure of amylopectin of waxy and nonwaxy rice. Agric Biol Chem. 1985; 49: 373-9.
• [45]  Jenkins PJ, Donald AM. Gelatinisation of starch: a combined SAXS/WAXS/DSC and SANS study. Carbohydr Res. 1998; 308: 133-47.