Research Papers
Development of a new selection method and quality improvement of sugary-1 rice mutants
2014 Volume 63 Issue 5 Pages 461-467
Details
2014 Volume 63 Issue 5 Pages 461-467
Brown rice of sugary-1 mutants has a wrinkled character because of the presence of phytoglycogen instead of starch in the inner part of the endosperm. Because the wrinkled phenotype was used as a sole selection marker for progeny of the sugary-1 strain, identification of mutant seeds with improved appearance is very difficult. We found that sugary-1 varieties contained not only phytoglycogen but also free glucose in the endosperm, and these were positively correlated. In the segregated F2 seeds that resulted from crossing Hokurikutou237 (sugary-1) and Koshihikari strains, glucose and phytoglycogen were also significantly correlated. Thus, we identified new sugary types with improved appearance from these progeny using glucose measurements. The F4 seeds of the improved strain had moderate phytoglycogen contents and seed germination characteristics. Native-PAGE showed that pullulanase activity in the improved strain increased in developing seeds compared with Hokurikutou237, although isoamylase activity was extremely low and similar to that in sugary-1 types. The new selection method in this study efficiently aids the development of improved sugary rice types that lack the wrinkled phenotype.
Starch is a water-insoluble granule consisting of crystalline and amorphous regions. It is predominantly composed of two α-glucan types, amylose and amylopectin. Amylose is predominantly linear and is composed almost entirely of α-1,4-glucosidic linkages, whereas amylopectin is a branched polymer of α-1,4-linked glucosyl residues with branched α-1,6-linkages. Biosynthesis of amylose and amylopectin is mediated by ADP-glucose pyrophosphorylases, starch synthases, starch branching enzymes and starch debranching enzymes (DBEs; Slattery et al. 2000, Smith et al. 1997).
DBEs catalyze the hydrolysis of α-1,6-glucan linkages of α-glucans. In higher plants, two types of DBEs with distinct substrate specificities have been identified, namely, the pullulanase type (EC 3.2.1.41) and the isoamylase type (EC 3.2.1.68; Doehlert and Knutson 1991). Although DBEs catalyze cleavage of α-1,6-glucan linkages of amylopectin during germination of cereal seed, analyses of sugary-1 mutants show that DBEs play a crucial role in starch biosynthesis. The sugary-1 mutants of maize and rice show a common endosperm phenotype, which is characterized by reduced starch content and accumulation of water soluble and highly branched polysaccharides such as phytoglycogen (James et al. 1995, Nakamura et al. 1996, Pan and Nelson 1984). Since the sugary-1 loci encodes genes for isoamylase 1 (Fujita et al. 1999, James et al. 1995, Yano et al. 1984), it is believed that formation of highly ordered amylopectin in vivo requires the actions not only of starch synthases and starch branching enzymes but also of DBEs. Pullulanase activity is closely related to the amount of starch accumulated in rice endosperm of allelic sugary-1 mutants. Moreover, studies suggest that pullulanse activity is also responsible for the formation of amylopectin-like α-glucans instead of phytoglycogen in the starch region of sugary-1 mutant lines (Kubo et al. 1999, Nakamura et al. 1997).
Several rice cultivars that have different starch composition and structure, including low amylose, high amylose, amylose extender and sugary type cultivars, have been bred in Japan and the structures and physical properties of their mutant starches were examined (Horibata et al. 2004, Inouchi 2010). The sugary type varieties, Ayunohikari and Hokurikutou237, which show the wrinkled phenotype of the parent sugary-1 mutants (EM-5) are intended as new materials for processing healthy foods in Japan because they accumulate approximately 3 times γ-aminobutyric acid (GABA) in germinated brown rice than in normal rice on a weigh basis (Miura et al. 2007). Furthermore the cooked sugary rice after polishing showed lowest digestibility among other cultivars with varied amylose contents (Arai et al. 2010). However, the wrinkled phenotype of the sugary type causes reduction in yield of brown rice and the seeds shape results in restriction of application of white rice due to unsuitable for rice polishing. Therefore, development of improved sugary rice is needed. In the absence of efficient methods, such as genomic markers, rice breeders use the wrinkled phenotype to select desirable strains. However, the identification of improved sugary-1 mutants by appearance is not possible. Thus, we have established a new efficient selection method to identify sugary-1 mutants with better appearance and we describe an improved sugary-1 mutant.
A total of 30 rice cultivars including two sugary types, Ayunohikari and Hokurikutou237, were used in this study. Other distinguishing phenotypes included waxy, low amylose (Sato et al. 2002), storage protein deficient (Nishimura et al. 2005), chalky type (Ashida et al. 2009, Kang et al. 2005) and giant embryo (Hong et al. 1996). These cultivars were grown and harvested in experimental fields at the NARO Agricultural Research Center. To analyze segregation of sugary type in this study, Hokurikutou237 was crossed with Koshihikari, which is normal type. Subsequently, F2 seeds derived from mature F1 plants were harvested in a greenhouse and were analyzed. The characterized F4 seeds in this study were harvested from one line of selected F3 with a fixed sugary and less wrinkled phenotype.
Measurement of glucose, phytoglycogen and GABA contentsGlucose and phytoglycogen contents were measured using a glucose assay kit (Glucose C2, Wako Pure Chemical Industries Ltd., Osaka, Japan) and a starch damage assay kit (Megazyme International Ireland, Ireland), respectively. The principle of both kits is the same, involving enzymatic glucose oxidation to yield hydrogen peroxide. In the presence of peroxidase, the hydrogen peroxide yields a red pigment in a quantitative oxidation condensation with phenol and 4-aminoantipyrine. Glucose concentration is proportional to the absorbance of the red color (Miwa et al. 1972). The starch damage assay kit contains α-amylase and amyloglucosidase to give complete degradation of starch-derived dextrins; in this study, phytoglycogen to glucose. To compare the glucose content in various rice cultivars, mature seeds were cut in half and the endosperm side without the embryo was used for the reaction. Water (50 μL) and the reaction solution in the kit (200 μL) were added to the half seed in a 96-well plate and were incubated at 37°C with mixing at 800 rpm for 30 min (Thermomixer Comfort; Eppendorf, New York, NY). After incubation, 200 μL of the mixture was transferred to another plate and the absorbance was measured using a microplate reader at 490 nm (Benchmark, Bio-Rad, Hercules, CA). The quantitative determination of glucose and phytoglycogen from Koshihikari, Hokurikutou237 and their progeny was performed as follows. Single seeds were crushed with a hammer, placed in a tube with 200 μL of water and were vortexed at room temperature for 1 h. The supernatant (50 μL) was centrifuged at 15,000 g for 10 min and was then placed in a 96-well plate with 200 μL of the reaction solution. Glucose content was then measured as described above. To measure phytoglycogen content, 50 μL of the supernatant was incubated with 50 μL of α-amylase solution (2.5 U) at 40°C for 20 min and the reaction was terminated by adding 400 μL of 0.2% (v/v) sulfuric acid solution. One tenth (50 μL) of the preparation was incubated with 50 μL of amyloglucosidase solution (1 U) at 40°C for 20 min and was continuously reacted with a color-producing reagent (2 mL) at 40°C for 20 min followed by measurement of absorbance at 510 nm. Phytoglycogen contents were calculated as a value subtracted free glucose content measured by a glucose assay kit.
Gamma-aminobutyric acid (GABA) in the samples was extracted after the removal protein by 2% sulfosalicylic acid solution and determined by the automatic amino acid analyzer (JLC-500/V2, JEOL Tokyo) according to the previous report (Suzuki et al. 2000). Samples were analyzed in triplicate.
Native-PAGE/Activity Staining of isoamylase and pullulanaseCrude enzymes were extracted from developing seeds. Forty kernels at the late-milky stage (about 15–20 days after flowering) were homogenized (20,000 rpm, 5 min) using a Polytron (Kinematica, AG., Switzerland) in 4 mL of grinding buffer containing 50 mM imidazole-HCl (pH 7.4), 8 mM MgCl2, 12.5% (v/v) glycerol and 50 mM 2-mercaptoethanol. Crude extracts were obtained by centrifugation at 15,000 g for 5 min at 4°C and then supernatants were filtered through a 0.45 μm filter. Native-PAGE was performed using 7.5% (w/v) acrylamide slab gels containing 0.15% (w/v) corn amylopectin (Sigma, St. Louis). Electrophoresis was carried out at 4°C with a constant current of 15 mA. After electrophoresis, the gel was rinsed and then incubated for 2 h at 30°C with 50 mL of 50 mM Citrate-Na2HPO4 buffer (pH 6.0) containing 50 mM 2-mercaptoethanol. The activities of starch DBEs were visualized by staining the gel with a solution containing 1% (w/v) KI and 0.1% (w/v) I2 (Kubo et al. 1999).
Pullulanase enzyme assayA reaction mixture containing 100 μL of 100 mM Mes-NaOH (pH 6.2), 20 mg/mL pullulanan (Nacalai Tesque, Kyoto, Japan) and 100 μL of suitably diluted enzyme solution from the above crude extract was incubated at 30°C for 20 min. The reaction was stopped by boiling for 5 min. Enzyme activities were measured as the increase in reducing power using dinitrosalicylic acid methods (Luchsinger and Cornesky 1962) with glucose as the standard. One unit of the enzyme activity was defined as 1 μmol of glucose equivalent/min/mg protein.
Seed properties and germination testsThe seeds of Hokurikutou237, Koshihikari and F4 cross of these parents were characterized. The seed shape (length, width, thickness and superficial area) and 1000-grain weight were measured 5 times using a grain quality analyzer (RGQI10; Satake Corporation, Tokyo, Japan). In the germination test for F4 seeds and parent seeds, approximately 20 brown rice seeds from each sample were immersed in water and incubated in the dark for 3 days at 30°C.
Statistical analysisMultiple sample comparisons were statistically analyzed using Excel Statistics functions (version 2010). Fisher’s least significant differences test was used to identify differences between means (p < 0.05).
Pedigrees of sugary-1 mutants showed that Ayunohikari and Hokurikutou237 were originated from the moderately mutated line EM-5 (Kubo et al. 1999) (Fig. 1A). Colorimetric determinations of glucose indicated that the original parent types Fukuhibiki, Itadaki and Kinmaze had much lower glucose contents than the current sugary-1 varieties (Fig. 1B). Glucose contents were compared between a total of 30 cultivars with normal, waxy, low amylose, storage protein mutant, chalky type and giant embryo phenotypes as well as the sugary type described above (Fig. 2). The results showed that Ayunohikari and Hokurikutou237 contained approximately 4 times more glucose than all of the other varieties and indicated that high glucose content is efficient method for the first screening of sugary-1 mutants and other glucose accumulated mutant lines.
Pedigree and glucose detection of rice sugary-1 mutants. (A) Mutagen treatments and sugary-1 mutants are indicated by squares and broken lines, respectively. (B) Simple visualization of glucose detection using half seeds of the sugary strains Ayunohikari, Hokurikutou237 and other parent strains.
Glucose contents of various rice cultivars. The 30 cultivars used in this study are classified as: a; normal, b; sugary type, c; waxy rice, d; low amylose type, e; storage protein deficient, f; chalky type and g; giant embryo. Error bars indicate the standard error from five independent experiments.
In this study, we initially confirmed that the starch damage assay kit could be used to determine phytoglycogen content. Preliminary experiments using commercial glycogen from oysters (Nacalai, Tesque Kyoto, Japan) showed that the absorbance at 510 nm increased in direct proportion to glycogen concentration (Supplemental Fig. 1).
To investigate the relationship between phytoglycogen and glucose contents in rice seeds, these contents were measured in 24 seeds of Koshihikari and Hokurikutou237. The assay showed that the glucose content of Koshihikari and Hokurikutou237 ranged from approximately 10 to 25 μg/seed and 25 to 60 μg/seed, respectively, whereas the phytoglycogen content of Koshihikari and Hokurikutou237 ranged from approximately 20 to 150 μg/seed and 200 to 700 μg/seed, respectively (Fig. 3A). Although the data from Hokurikutou237 were widely spread compared with those of Koshihikari, phytoglycogen and glucose contents were well correlated (r = 0.936). To confirm availability as a selection method, we crossed the Koshihikari and Hokurikutou237 cultivars and obtained F1 seeds. Segregation analysis was carried out in the F2 seeds (a total of 202 seeds) from the mature F1 plant (Fig. 3B). Although the data showing glucose and phytoglycogen contents from the F2 seeds were distributed more widely than those of the parent seeds, the distribution still indicated a considerable correlation (r = 0.804). Furthermore, the segregation ratio among the F2 seeds showed a good fit to the expected 3 : 1 ratio (observed values 159 : 43) with a p value of 0.22, which corresponded to previous reports showing that the sugary phenotype was controlled by a single recessive gene (Yano et al. 1984). In addition, the ratio between non-wrinkled and wrinkled seeds in F2 population also showed the expected values 161 : 41 with a p value of 0.12.
Relationship between glucose and phytoglycogen content in F2 and parent seeds. (A) Parent lines; open and closed dots indicate the contents of Hokurikutou237 and Koshihikari seeds, respectively. Solid and broken circles enclose data points for normal and sugary-1 mutant rice, respectively. (B) F2 seeds (cross between Hokurikutou237 and Koshihikari); the broken line shows the expected border between normal and sugary-1 mutant rice using parent data as a reference.
During segregation analysis of the described F2 seeds, there were several seeds of the sugary type that were less wrinkled despite their phytoglycogen contents. We cultivated these F2 plants and screened lines producing F3 seeds with a fixed sugary, but less wrinkled phenotype. Appearances of the F4 seeds derived from a selected F3 line and their parents are shown in Fig. 4. The F4 seeds had a moderately wrinkled phenotype compared with Hokurikutou237 and had increased opacity (data not shown). Iodine staining of starch indicated that the non-stained phytoglycogen region in the center of the F4 seed endosperm was smaller than that of the Hokurikutou237 cultivar (Fig. 4). The average phytoglycogen content of the F4 seeds (339 μg/seed) was also lower than that of the Hokurikutou237 seeds (433 μg/seed) but much higher that of the Koshihikari seeds (56.6 μg/seed). Similarly, glucose contents of the F4 seeds (29.7 μg/seed) was a little lower than that of the Hokurikutou237 seeds (31.6 μg/seed) but was higher that of the Koshihikari seeds (13.0 μg/seed). Although the appearance of brown rice was improved, 1000-grain weight of the F4 seeds (15.8 g) showed little difference to Hokurikutou237 seeds (15.2 g) and was lower than that of the Koshihikari seeds (20.2 g). On the other hand, GABA contents of the F4 seeds (48.3 mg/100 g seeds) was a little lower than that of the Hokurikutou237 seeds (53.8 mg/100 g seeds) but was higher that of the Koshihikari seeds (21.1 mg/100 g seeds) and indicates that F4 seeds still maintains health benefit.
Seed morphology of the sugary F4 and parent seeds. Whole brown rice is shown in the upper panels and the cross-sections of mature seeds stained with iodine solution are shown in the lower panels. Average glucose, phytoglycogen and GABA contents and 1000-grain weight are shown in the table. Data are presented as a percentage of the Koshihikari control in parenthesis. The letters a, b and c indicate significant differences (p < 0.05).
To determine changes in the debranching enzymes, isoamylase (ISA) and pullulanase (PUL), in the developing seeds, native-PAGE/DBE activity staining of the crude extraction from the developing endosperms of the Hokurikutou237, Koshihikari and F4 seeds was performed (Fig. 5). Both debranching enzymes were detected as blue bands on the native-PAGE gel containing corn amylopectin stained with an iodine solution. Although ISA activity was detected as a major and a minor band with low mobility in the Koshihikari seeds, no such blue band was detected in the Hokurikutou237 and F4 seeds, confirming that the selected mutant line with improved appearance had inherited the ISA deficiency of the EM-5 mutant. The activity of PUL in the Koshihikari seeds was also detected as a single major blue band with greater mobility than ISA and in Hokurikutou237 slight enzyme activity was detected. Interestingly, the PUL activity of the F4 seeds was about half that of the Koshihikari seeds. Using pullulane as a substrate, the specific PUL activity in the Koshihikari, F4 and Hokurikutou237 seeds was 1.19, 0.72 and 0.35 U/mg, respectively.
Activities of debranching enzymes in the endosperm at the developing stage of the sugary F4 and parent seeds. Crude extracts from the endosperm were applied (15 μL per lane) to a 7.5% polyacrylamide gel. Arrows indicate bands corresponding to isoamylase (ISA), pullulanase (PUL) and phosphorylase (PHO).
Seeds were germinated at 30°C for 3 days. All tested brown rice seeds of Hokurikutou237 germinated and few of the Koshihikari strain sprouted. As a hybrid of these two varieties, about half of the mature F4 seeds were germinated in the same conditions (Fig. 6).
Seed germination tests. (A), (B) and (C) show the sprouting of Koshihikari, F4 and Hokurikutou237 seeds, respectively.
The developing of DNA marker is generally most sufficient and straightforward for mutant selection not only sugary-1. However, it is difficult to build DNA marker of sugary-1 at present because the mutated point of the isoamylase1 locus remain to be delineated in spite of many researches. Although the measurement of phytoglycogen in seeds also is a straightforward method for identifying the sugary phenotype, in contrast with the glucose assay, the requirement of α-amylase and amyloglucosidase enzyme reaction steps for degradation of phytoglycogen to glucose increases cost and time. Hence the easy method is required for sugary-1 selection.
Comparison of the glucose contents among various cultivars showed that only the sugary types, Ayunohikari and Hokurikutou237, contained high levels of glucose (Fig. 2). The cultivars examined in this study were grown in different regions of the country at different times. As the composition of rice fluctuates depending on environmental conditions, these observations suggest measurement of glucose content is appropriate for selection of sugary progeny of various parent varieties and is impervious to the environment. Segregation analysis showed that the distinction between sugary and non-sugary types was obscure compared with the parent lines and several F2 seeds had little phytoglycogen despite their high glucose content (Fig. 3B). However, using a binomial glucose threshold of 30 μg/seed, we selected sugary-1 mutants with more than 95% confidence. Hence these data provide a simple method for estimating phytoglycogen content and screening sugary-1 mutants that lack the wrinkled phenotype.
We obtained a new sugary type with a moderately wrinkled phenotype by crossing the Koshihikari and Hokurikutou237 strains. The F4 seeds of the new strain also showed well correlation between phytoglycogen and glucose (r = 0.883) similar to Hokurikutou237, indicated that our selection method could fix sugary-1 on F4 seeds (Supplemental Fig. 2). The F4 seeds showed little difference in 1000-grain weight compared with the sugary parent type Hokurikutou237 and had improved brown rice appearance (Fig. 4). Despite this, grain thickness of the F4 seeds was increased (1.66 mm) in comparison with the Hokurikutou237 grains (1.57 mm; Fig. 4). However, the length and width of the F4 seeds was smaller than that of the parent grains, resulting in decreased surface area (Supplemental Table 1). Although there was little difference in the weight of the F4 seeds, recovery from the wrinkled phenotype allowed polishing of sugary type brown rice for the first time (data not shown).
Native-PAGE/DBE staining (Fig. 5) and PUL assays showed that PUL activity in developing F4 seeds increased in comparison with the sugary parent without changing ISA activity. These data suggest that PUL activity in the developing F4 seeds may contribute to normal starch synthesis, resulting in decreased phytoglycogen and improved appearance. In support of these observations, previous reports show that PUL activities are reduced in sugary-1 endosperms and PUL protein and enzyme activity is closely related to the magnitude of the starch region (Kubo et al. 1999, Nakamura et al. 1997). However, the effects of other enzymes of starch biosynthesis, such as ADP-glucose pyrophosphorylase, starch synthase and starch branching enzyme, on the phenotype of the F4 seeds is unclear. In contrast, the activities of these enzymes in the original sugary-1 mutant EM-5 and in PUL deficient mutant lines were basically unaffected (Fujita et al. 2009, Kubo et al. 1999). Interestingly, previous reports showed that transcription of PUL mRNA in sugary-1 and maize su1 mutants (Beatty et al. 1999) were not different from those of the wild type. However, the protein expression and enzyme activity of PUL was decreased in sugary-1 mutant lines. These results imply that PUL might be susceptible to degradation by proteases (Fujita et al. 2009). Although the explanation for increased PUL activity in the developing F4 seeds remains unclear, one possibility may be changing the background except for ISA from Hokurikutou237 to Koshihikari such as proteases or chaperone proteins that influence PUL stability. Our new selection method may provide an unprecedented tool for separating the effects of ISA and PUL.
For agronomic applications, the regulation of seed dormancy is important for preventing preharvest sprouting and achieving uniform germination. The sugary type cultivars, Ayunohikari and Hokurikutou237, are easily germinated and need to be harvested early in seed maturing, approximately 25–30 days after heading by contrast to 35–40 days of Koshihikari (Miura et al. 2007). The F4 strain showed inferior germination to Hokurikutou237 due to its phytoglycogen content. This observation is explained by the availability of energy from the water soluble polysaccharide phytoglycogen as compared with that from starch.
We provide a simple method for screening improved sugary rice mutants and have developed a new sugary line that is moderately wrinkled, and improved germination characteristics that result from recovery of PUL activity. Future studies of starch and phytoglycogen contents of new sugary rice strains will lead to further agronomic improvements.
We thank many rice breeders in NARO for providing various rice seeds. The authors would like to thank Enago (www.enago.jp) for the English language review.
