Amyloplast Membrane Protein SUBSTANDARD STARCH GRAIN6 Controls Starch Grain Size in Rice Endosperm1 Ryo Matsushima*, Masahiko Maekawa, Miyako Kusano, Katsura Tomita, Hideki Kondo, Hideki Nishimura, Naoko Crofts, Naoko Fujita, and Wataru Sakamoto Institute of Plant Science and Resources, Okayama University, Kurashiki 710–0046, Japan (R.M., M.M., H.K., H.N., W.S.); Graduate School of Life and Environmental Sciences, University of Tsukuba, Tsukuba, Ibaraki 305–8572, Japan (M.K.); RIKEN Center for Sustainable Resource Science, Yokohama, Kanagawa 230–0045, Japan (M.K.); Fukui Agricultural Experiment Station, Fukui 918–8215, Japan (K.T.); and Department of Biological Production, Akita Prefectural University, Akita 010–0195, Japan (N.C., N.F.)

Starch is a biologically and commercially important polymer of glucose. Starch is organized into starch grains (SGs) inside amyloplasts. The SG size differs depending on the plant species and is one of the most important factors for industrial applications of starch. There is limited information on genetic factors regulating SG sizes. In this study, we report the rice (Oryza sativa) mutant substandard starch grain6 (ssg6), which develops enlarged SGs in endosperm. Enlarged SGs are observed starting at 3 d after flowering. During endosperm development, a number of smaller SGs appear and coexist with enlarged SGs in the same cells. The ssg6 mutation also affects SG morphologies in pollen. The SSG6 gene was identified by map-based cloning and microarray analysis. SSG6 encodes a protein homologous to aminotransferase. SSG6 differs from other rice homologs in that it has a transmembrane domain. SSG6-green fluorescent protein is localized in the amyloplast membrane surrounding SGs in rice endosperm, pollen, and pericarp. The results of this study suggest that SSG6 is a novel protein that controls SG size. SSG6 will be a useful molecular tool for future starch breeding and applications.

Starch is an important plant-derived Glc polymer that is widely used as food and in manufacturing applications involving nonfood products. Starch is water insoluble and osmotically inactive. These properties make starch a suitable molecule for long-term carbohydrate storage in seeds and tubers. Higher plant cells contain terminally differentiated plastids called amyloplasts, which is the organelle involved in starch synthesis and storage in endosperm and tubers (Sakamoto et al., 2008). Starch is organized as transparent grains (starch grains [SGs]) in amyloplasts (Buléon et al., 1998; Hancock and Tarbet, 2000). SGs are easily visualized using iodine solution and can be clearly observed with a normal light microscope (Matsushima et al., 2010). Cereal endosperm accumulates high levels of starch, which fill most of the amyloplast intracellular space. 1 This work was supported by the Ministry of Education, Culture, Sports, Science, and Technology (Grant-in-Aid for Scientific Research no. 26450004 to R.M.), the Japan Advanced Plant Science Network, and the Information Center of Particle Technology. * Address correspondence to [email protected]. The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Ryo Matsushima ([email protected]). R.M. designed the experiments; R.M., M.M., M.K., K.T., H.K., H.N., N.C., and N.F. performed the experiments; R.M., M.M., and W.S. discussed the results; R.M. conceived the project and wrote the article. www.plantphysiol.org/cgi/doi/10.1104/pp.15.01811

Therefore, the amyloplast volume is considered as approximately equivalent to the SG volume (Matsushima et al., 2014). Rice (Oryza sativa) endosperm SGs are normally 10 to 20 mm in diameter (Matsushima et al., 2010). Each amyloplast contains a single SG that is organized from the assembly of several dozen smaller starch granules. Each starch granule is a sharp-edged polyhedron with a typical diameter of 3 to 8 mm. This type of SG is called a compound SG (Tateoka, 1962). For compound SGs, starch granules are assembled but not fused to form a single SG, which is easily separated by conventional purification procedures. By contrast, a simple SG contains a single starch granule (Tateoka, 1962). In this case, both terms are used equally. Simple SGs are produced in several important crops such as maize (Zea mays), sorghum (Sorghum bicolor), barley (Hordeum vulgare), and wheat (Triticum aestivum; Tateoka, 1962; Matsushima et al., 2010, 2013). SG sizes in cereal endosperm are diverse. Maize and sorghum SGs have a uniform size distribution of approximately 10 mm in diameter (Jane et al., 1994; Matsushima et al., 2010; Ai et al., 2011). In barley and wheat, SGs of two discrete size classes (approximately 15225 mm and less than 10 mm) coexist in the same cells (Evers, 1973; French, 1984; Jane et al., 1994; Matsushima et al., 2010, 2013). The size of starch granules is one of the most important characteristics for industrial applications (Lindeboom et al., 2004). Small starch granules can replace fat in food applications, because aqueous dispersions of small

Plant PhysiologyÒ, March 2016, Vol. 170, pp. 1445–1459, www.plantphysiol.org Ó 2016 American Society of Plant Biologists. All Rights Reserved.

1445

Matsushima et al.

Figure 1. Enlarged SGs of mature ssg6 endosperm. A, Wild-type cv Koshihikari (WT) seeds. Bar = 1 mm. B, ssg6 seeds. Bar = 1 mm. C, Quantification of wild-type and ssg6 seed sizes (n = 30 each). D, One-hundred seed weight of the wild type and ssg6 (n = 16 and 15, respectively). E and G, Iodine-stained thin sections of wild-type endosperm at low and high magnification, respectively. Bars = 10 mm. F and H, Iodine-stained thin sections of ssg6 endosperm at low and high magnification, respectively. Bars = 10 mm. I, Quantification of the areas occupied by SGs in sections of different genotypes (n $ 30 each). J, Quantification of the starch amount in mature seeds expressed as the percentage of seed weight (n = 6 each). Data are given as means 6 SD. Statistical comparisons were performed using Welch’s t test; all data were compared with the wild type (**, P , 0.01; *, P , 0.05; and ns, not significant at P = 0.05).

starch granules show fat-mimetic properties (Malinski et al., 2003). The larger starch granules of maize and cassava (Manihot esculenta) improve the final starch yield after wet-milling purification (Gutiérrez et al., 2002). In the case of simple SGs, the size of SGs is equal to the size of starch granules. Therefore, manipulation of the sizes of SGs and starch granules is a molecular target for bioengineering programs. SG size can be reduced in transgenic plants and genetic mutants by down-regulating several starch synthetic enzymes

(Gutiérrez et al., 2002; Bustos et al., 2004; Ji et al., 2004; Stahl et al., 2004; Matsushima et al., 2010; Fujita, 2014). By contrast, our understanding of the genetic tools, biosynthetic enzymes, and plant materials that can be utilized to enlarge SGs is limited. Recent work identified a rice mutant that develops enlarged SGs; this mutant has been named substandard starch grain4 (ssg4; Matsushima et al., 2014). The enlarged SGs are observed in starch-accumulating tissues of ssg4, including endosperm, pollen, root caps,

Table I. Effects of the ssg6 mutation on gelatinization properties of endosperm starch determined by differential scanning calorimetry Gelatinization properties of starch in ssg6 seeds were analyzed by differential scanning calorimetry. Values are means 6 SD of nine independent determinations. TO, TP, and TC are onset, peak, and conclusion temperatures, respectively. DH is gelatinization enthalpy of starch. Asterisks indicate significance at P , 0.01.

1446

Plant

TO

TP

TC

DH

Wild-type cv Koshihikari ssg6

˚C 54.9 6 0.7 56.3 6 1.7

63.9 6 0.6 65.8 6 0.9

71.2 6 0.2 72.9 6 1.0

mJ mg21 8.6 6 1.3 9.3 6 1.1 Plant Physiol. Vol. 170, 2016

Rice Mutant with Enlarged Starch Grains

and young pericarp. Chloroplasts in young ssg4 leaves also are enlarged. SSG4 encodes an amyloplastlocalized large protein with a domain of unknown function (DUF490). SSG4 homologs are conserved from bacteria to higher plants; however, the exact molecular function of SSG4 is unknown. In this study, we report the identification of another rice mutant (ssg6) that develops enlarged SGs in

endosperm. We characterize the ssg6 mutation and identify the responsible gene. SSG6 encodes a protein homologous to aminotransferase. SSG6 is localized at the amyloplast membrane and is a novel factor that influences SG size. We also determined that ssg4 and ssg6 mutations act synergistically in pollens. SSG6 will be a useful molecular target for future starch breeding and starch biotechnology programs.

Figure 2. Developing SGs in maturing endosperm. A to C, Wild-type cv Koshihikari (WT) seeds at 3, 5, and 7 DAF, respectively. Bars = 1 mm. D to F, ssg6 seeds at 3, 5, and 7 DAF, respectively. Bars = 1 mm. G to I, Iodine-stained thin sections of wild-type endosperm at 3, 5, and 7 DAF, respectively. Bars = 20 mm. J to L, Iodine-stained thin sections of ssg6 endosperm at 3, 5, and 7 DAF, respectively. Bars = 20 mm. M, Quantification of the areas occupied by SGs at 3, 5, and 7 DAF (n $ 11 each). N and O, Higher magnifications of iodinestained thin sections of wild-type and ssg6 endosperm at 7 DAF, respectively. Bars = 10 mm. P, Kernel density plot for the areas occupied by SGs at 7 DAF. The wild type and ssg6 are indicated by red and black lines, respectively (n = 300 and 299, respectively). Q, Quantification of the numbers of SGs per 5,833 mm2 at 7 DAF. Data are given as means 6 SD. Statistical comparisons were performed by Welch’s t test; all data were compared with the wild type (**, P , 0.01; and ns, not significant at P = 0.05). Plant Physiol. Vol. 170, 2016

1447

Matsushima et al.

Figure 3. SGs in pollen and chloroplasts in third leaves. A and B, Iodine-stained pollen of wild-type cv Koshihikari (WT) and ssg6, respectively. Bars = 10 mm. C and D, Released SGs from squashed pollen of the wild type and ssg6, respectively. Bars = 10 mm. E, Quantification of the areas occupied by SGs in pollen (n = 40 each). F and G, Third leaves of the wild type and ssg6, respectively. Bars = 1 mm. H and I, Thin sections of the third leaves of the wild type and ssg6 were double stained with Methylene Blue and Basic Fuchsin. Bars = 10 mm. J and K, Magnified images of H and I. Bars = 10 mm. L, Quantification of the areas occupied by chloroplasts in third leaves (n = 16 each). The quantification was based on three and five TEM sections of the wild type and ssg6, respectively. Data are given as means 6 SD. Statistical comparisons were performed by Welch’s t test (**, P , 0.01). M and N, TEM images of chloroplasts of the wild type and ssg6, respectively. Bars = 1 mm.

1448

Plant Physiol. Vol. 170, 2016

Rice Mutant with Enlarged Starch Grains

Table II. Segregation of ssg6 pollen phenotypes in F1 plants Mature anthers from the F1 hybrid between ssg6 and wild-type cv Koshihikari were disrupted with forceps in dilute Lugol solution on a glass slide to obtain iodine-stained pollen grains. The released pollen were subsequently examined with the microscope. Parental Genotype

ssg62/+

No. of Wild-Type Pollen

No. of ssg6 Pollen

Total

Percentage of ssg6 Pollen

31

24

55

43.6

RESULTS Enlarged SGs in Mature ssg6 Endosperm

We previously developed a rapid method to observe SGs in rice endosperm (Matsushima et al., 2010). Using this method, we performed a genetic screen of a mutagenized cv Nipponbare population and isolated mutants with defective SGs (Matsushima et al., 2010). We repeated these experiments using cv Koshihikari, which is an elite variety in Japan. In this study, we performed the same genetic screen using a mutagenized cv Koshihikari population and isolated ssg6. The ssg6 seed morphology was slightly chalky compared with wild-type cv Koshihikari seeds (Fig. 1, A and B). Polished ssg6 grains were chalky (Supplemental Fig. S1A). The ssg6 grain size was slightly smaller than that of the wild type, especially with respect to grain width and depth (Fig. 1C). The ssg6 grain weight was less than that of the wild type (Fig. 1D). Iodine-stained thin sections of mature ssg6 endosperm clearly showed enlarged SGs (Fig. 1, E–H). Quantification of SG areas in thin sections showed that ssg6 SGs were more than 2 times larger than those in the wild type (Fig. 1I). Next, we quantified the number and size of starch granules in SGs of mature endosperm. The number of starch granules was increased in enlarged SGs of ssg6 compared with wild-type SGs (Supplemental Fig. S1B). In contrast, the size of starch granules decreased slightly in ssg6 (Supplemental Fig. S1C). Therefore, the increased size of ssg6 SGs was mainly attributed by the increased numbers of starch granules. Endosperm is a triploid tissue. Therefore, endosperm has four possible genotypes at one gene locus (i.e. AAA, AAa, Aaa, and aaa). We performed reciprocal crosses to obtain two distinct heterozygous seeds of SSG6SSG6ssg6 and SSG6ssg6ssg6. The SG sizes of heterozygous seeds and wild-type seeds did not differ significantly (Fig. 1I). This indicated that one copy of the wild-type allele supplied from the male gametophyte was sufficient for the formation of normal-sized SGs. We examined starch accumulation in ssg6 seeds. The total amount of starch as a percentage of weight was not significantly different between ssg6 and wild-type seeds (Fig. 1J). Amylopectin chain-length distribution and amylopectin gelatinization temperature were similar in ssg6 and wild-type starch (Table I; Supplemental Fig. S1D). The amylose content was slightly lower in ssg6 starch than in wild-type starch (Supplemental Fig. S1E). We analyzed soluble sugars in ssg6 and wild-type Plant Physiol. Vol. 170, 2016

seeds using gas chromatography-time-of-flight-mass spectrometry. Glc, Fru, and Suc levels were significantly higher in ssg6 than in wild-type seeds (Supplemental Fig. S2). Man, trehalose, Xyl, and raffinose levels were not significantly different in ssg6 and wild-type seeds (Supplemental Fig. S2). Dynamic Behavior of SGs in Developing Endosperm

We investigated when enlarged SGs developed in the ssg6 mutant by analyzing developing seeds at 3, 5, and 7 d after flowering (DAF). From 3 to 7 DAF, seed enlargement was essentially the same in ssg6 and the wild type (Fig. 2, A–F), but SG sizes differed (Fig. 2, G–L). At 3 DAF, most SGs in ssg6 endosperm were larger than those in the wild type (Fig. 2, G and J). We also used transmission electron microscopy (TEM) to assess ssg6 endosperm at 3 DAF (Supplemental Fig. S3). The ssg6 SG morphologies in TEM images were spherical and large, similar to the iodine-stained SGs. The ssg6 enlarged SGs occupied an area more than 10-fold larger than that of normal-sized wild-type SGs (Fig. 2M). At 7 DAF, the area occupied by SGs was more than 2.5-fold larger in ssg6 than in the wild type (Fig. 2M). The quantification does not include the very small SGs, in order to compare only the enlarged SGs. Meanwhile, the number of small SGs increased and coexisted with the enlarged SGs in the same cells during endosperm development in ssg6. At 7 DAF, the outlines of both small and enlarged SGs were sharp (Fig. 2, N and O). Therefore, we quantified the sizes of both small and enlarged SGs in 7-DAF endosperm. At 7 DAF, wild-type endosperm contained SGs with uniform sizes (Fig. 2N). By contrast, ssg6 endosperm contained a number of smaller SGs in addition to the enlarged SGs (Fig. 2O). Kernel density estimation indicated that ssg6 contained primarily small SGs less than 50 mm2 (Fig. 2P). This estimation was determined using two-dimensional sections; therefore, the method may underestimate SG sizes and overestimate small SG densities due to SGs oriented in a nonsagittal plane. However, small SGs appear to be more abundant in ssg6 than in the wild type (Fig. 2, N and O). In ssg6 endosperm, enlarged SGs with areas greater than 400 mm2 were not as abundant as small SGs. In wild-type endosperm, the majority of SGs had sizes of 100 to 200 mm2. The total numbers of SGs per area were approximately the same in ssg6 and the wild type (Fig. 2Q). At 3 and 5 DAF, the small SGs of ssg6

Table III. Segregation of ssg6 seeds in the F2 population F2 seeds were obtained from the cross between ssg6 and wild-type cv Koshihikari. Endosperm thin sections were prepared from 129 F2 seeds. SG size was examined with the microscope. Parental Genotype

ssg62/+

No. of ssg6 Seeds

No. of Wild-Type Seeds

Total

x 2 Value (P Value) for 1:3 Segregation

30

99

129

0.21 (0.65) 1449

Matsushima et al.

were not stained clearly with iodine and did not have clear outlines. That made it difficult to quantify their sizes. We also investigated the protein accumulation of starch biosynthetic enzymes in developing seeds (Supplemental Fig. S4). Accumulation levels of starch synthases, branching enzymes, phosphorylase, and pullulanase did not change between wild-type and ssg6 developing seeds.

Pollen SGs and Leaf Chloroplasts

Pollen, root cap, and pericarp also accumulate high levels of starch in rice. The ssg6 pollen developed spherical SGs, whereas wild-type pollen developed

rod-like SGs (Fig. 3, A–D). Areas occupied by SGs were slightly less in ssg6 than in the wild type (Fig. 3E). When ssg6 was crossed with the wild type, approximately half of the F1 pollen displayed ssg6 phenotypes (Table II). This indicates that ssg6 behaves in a gametophytic manner in pollen. The SGs in root cap and pericarp were similar in ssg6 and the wild type (Supplemental Fig. S5). A variegated phenotype was observed in the third leaves of ssg6 mutants (Fig. 3, F and G). Therefore, we speculated that chloroplasts were affected by the ssg6 mutation. To visualize chloroplasts, thin sections of third leaves from young plants were double stained with Methylene Blue and Basic Fuchsin (Fig. 3, H–K). Wild-type chloroplasts displayed lens-like shapes, whereas ssg6 chloroplasts were more spherical.

Figure 4. Map-based cloning of the SSG6 gene. A, Fine-mapping of the SSG6 locus on chromosome 6. Phenotypes and genotypes of 205 F2 progeny (410 chromosomes) were analyzed. The numbers of recombinations that occurred between SSG6 and the molecular markers are indicated. The SSG6 locus was mapped to a 152-kb region between two molecular markers (RM588 and RM190). This region contains 29 open reading frames (boxes). The ssg6 mutant has a mutation in Os06g0130400 (gray box). The position (1,629,68521,633,530) is based on Os-Nipponbare-Reference-IRGSP-1.0. B, Schematic representation of the exon and intron organization of Os06g0130400 and its complementary DNA (cDNA) obtained from RACE analysis. The deduced protein structure is shown. Numbers in parentheses are the positions of chromosome 6 based on Os-Nipponbare-Reference-IRGSP-1.0. The ssg6 mutant has a base pair change (C-to-T) at position 1,630,075 (red arrow). SSG6 encodes a protein containing 542 amino acid (a.a.) residues. A putative transmembrane domain and the aminotransferase class I and II domain (Aminotran_1_2 domain) are indicated by red and yellow boxes, respectively. The ssg6 mutation introduces a premature stop codon at amino acid 63 of SSG6 (red arrow). C, Phylogenic relationships of SSG6 homologs. Arabidopsis and rice homologs are indicated by black and blue labels, respectively. Previous studies have shown that ACS10 and ACS12 are aminotransferases, whereas ACS2, ACS4, ACS5, ACS6, ACS7, ACS8, ACS9, and ACS11 have the activity of ACS. The sequence names are from The Arabidopsis Information Resource (TAIR) and RAP databases. Bootstrap values from 1,000 trials are indicated. The 0.1 scale shows substitution distance. D, Prediction of transmembrane domains in SSG6 and homologs. Predictions were performed using TMHMM Server version 2.0. SSG6, ACS10, and ACS12 had putative transmembrane domains at the N termini. 1450

Plant Physiol. Vol. 170, 2016

Rice Mutant with Enlarged Starch Grains

Chloroplast areas were approximately 3-fold larger in ssg6 than in the wild type (Fig. 3L). These results indicate that the ssg6 mutation affects chloroplast size. We used TEM to observe ssg6 chloroplast ultrastructures, starch granules, and grana stacks. The results showed that starch granule sizes and grana stacks were similar in ssg6 and wild-type chloroplasts (Fig. 3, M and N). Based on the TEM images, we quantified the number of chloroplasts per cell and the number of starch granules per chloroplast (Supplemental Fig. S6, A and B). In the ssg6 third leaves, the number of chloroplasts in one cell were decreased compared with the wild type (Supplemental Fig. S6A). The enlarged chloroplasts in ssg6 third leaves tend to contain many more starch granules compared with the wild-type chloroplasts (Supplemental Fig. S6B). This is consistent with the previous observation that enlarged chloroplasts in Arabidopsis (Arabidopsis thaliana) plastid division mutants contained increased numbers of starch granules (Crumpton-Taylor et al., 2012). The ssg6 flag leaves were not variegated (Supplemental Fig. S7, A and B). The flag leaf chloroplasts were slightly larger in ssg6 than in the wild type but did not differ to the extent of chloroplasts in the third leaves of ssg6 and the wild type (Supplemental Fig. S7, C–G). TEM analysis of chloroplast ultrastructure in flag leaves indicated that there were no significant differences between ssg6 and the wild type (Supplemental Fig. S7, H and I). As in the third leaf, the number of chloroplasts per cell was decreased and the number of starch granules per chloroplast was increased in ssg6 flag leaves (Supplemental Fig. S6, C and D).

transition at position 1,630,075 (Os-Nipponbare-ReferenceIRGSP-1.0-based position). The C-to-T transition was consistent with an ethyl methanesulfonate-induced mutation. Os06g0130400 encoded a protein containing 542 amino acid residues (Fig. 4B). The mutation in ssg6 introduced a premature stop codon at amino acid position 63 (Fig. 4B; Supplemental Fig. S9). The Os06g0130400 protein was predicted to have an

Identification of the SSG6 Gene

The ssg6 phenotype in endosperm tissue was completely penetrant in ssg6-selfed progeny and segregated as a single recessive allele in F2 progeny (Table III). We identified the SSG6 gene using conventional map-based cloning. We mapped the ssg6 mutation within a 152-kb region on chromosome 6 (Fig. 4A). Twenty-nine open reading frames were expected in this region according to the Rice Annotation Project (RAP) database (http:// rapdb.dna.affrc.go.jp/). To narrow down the candidate genes, we performed DNA microarray analysis to investigate gene expression in 4-DAF seeds. We expected the expression of the mutated gene to be reduced in ssg6 mutants. Out of the 29 candidate genes, 20 genes had at least one probe on the Agilent Rice Oligo DNA Microarray 44K that we used. Of these, a single probe (Os06g0130400|mRNA|AK065212|CDS+39UTR) showed significantly reduced signal intensity in the ssg6 mutant compared with that of the wild type (Supplemental Fig. S8). This indicated that Os06g0130400 gene expression was reduced in ssg6. Next, we determined the nucleotide sequence of Os06g0130400 in ssg6 and the wild type. We identified a base change in the Os06g0130400 gene of ssg6 (Supplemental Fig. S9); the mutant carried a C-to-T Plant Physiol. Vol. 170, 2016

Figure 5. Complementation of the ssg6 mutant by SSG6 cDNA. SSG6 cDNA was expressed under the control of the maize UBIQUITIN1 promoter in the ssg6 mutant background. Two independent transgenic plants (Ubi:SSG6cDNA/ssg6 #1 and #4) were examined using the following tissues: mature endosperm (A–D) and pollen (E–H). A and B, Low magnification. C and D, High magnification. E and F, Images of whole-mount pollen grains. G and H, SGs that were released from squashed pollen grains. Bars = 10 mm. 1451

Matsushima et al.

aminotransferase class I and II domain (Aminotran_1_2 domain) according to the Pfam database (Fig. 4B; Supplemental Fig. S9). We performed a complementation test using the maize UBIQUITIN1 promoter to express Os06g0130400 cDNA. The cDNA sequence was confirmed by 59 and 39 RACE experiments. The cDNA clone was introduced into the ssg6 mutant, and transgenic plants homozygous for the transgene were isolated and named UBi:SSG6cDNA/ ssg6. SG sizes and morphologies in UBi:SSG6cDNA/ssg6 endosperm and pollen were very similar to those in the wild type (Fig. 5). This indicates that SG phenotypes in endosperm and pollen were completely rescued by the transgene. We conclude that Os06g0130400 is the gene responsible for the ssg6 mutation. SSG6 (Os06g0130400) shows homology with 1aminocyclopropane-1-carboxylate synthase (ACS) and was named OsACS6 according to the RAP database. Arabidopsis and rice genomes contain 12 and six annotated ACS genes, respectively (Yamagami et al., 2003; Iwai et al., 2006). In Arabidopsis, ACS1 was enzymatically inactive and ACS3 was a pseudogene, whereas

ACS2, ACS4, ACS5, ACS6, ACS7, ACS8, ACS9, and ACS11 encode enzymatically active ACS proteins (Yamagami et al., 2003). ACS10 and ACS12 were shown to have aminotransferase but not ACS activity (Yamagami et al., 2003). SSG6 (OsACS6) was phylogenetically related to ACS10 and ACS12 (Fig. 4C; Supplemental Fig. S10). In rice, other ACS homologs (OsACS12OsACS5) belonged to ACS. The transmembrane prediction program TMHMM Server version 2.0 (http://www.cbs.dtu.dk/services/ TMHMM-2.0/) predicted one transmembrane domain in the SSG6 protein (Fig. 4D). One transmembrane domain also was predicted for Arabidopsis ACS10 and ACS12 (Fig. 4D). By contrast, no transmembrane domains were predicted in rice ACSs (OsACS12OsACS5).

Subcellular Localization of the SSG6 Protein

SSG6 contains one putative transmembrane domain (Fig. 4D). Therefore, SSG6 may be localized to an intracellular membrane. To analyze the subcellular

Figure 6. Amyloplast membrane localization of SSG6-GFP in various tissues. Confocal and differential interference contrast (DIC) images of SSG6-GFP transgenic plants are shown. A to C, Whole-mount images of pollen. Bar = 5 mm. D to F, Higher magnification images of amyloplasts in pollen. Bar = 5 mm. G to I, Endosperm sections obtained by vibratome sectioning. Bar = 5 mm. J to L, Pericarp sections obtained by vibratome sectioning. Bar = 5 mm.

1452

Plant Physiol. Vol. 170, 2016

Rice Mutant with Enlarged Starch Grains

Figure 7. Agronomic traits of ssg6 plants. A, Wild-type cv Koshihikari (WT) and ssg6 grown under natural paddy field conditions. Bar = 10 cm. B, Duration to heading. C, Culm length. D, Panicle length. E, Number of panicles. F, Number of spikelets per panicle. G, Number of sterile spikelets per panicle. H, Total panicle weight. n = 16 plants for the wild type and 15 plants for BC3F2 homozygous ssg6. Data are given as means 6 SD. Statistical comparisons were performed by Welch’s t test (**, P , 0.01; and ns, not significant at P = 0.05).

localization of SSG6, we generated a construct containing the full-length SSG6 cDNA fused to GFP (SSG6GFP). SSG6-GFP was transiently expressed in Nicotiana benthamiana leaves, and SSG6-GFP fluorescence was detected surrounding chlorophyll autofluorescence (Supplemental Fig. S11). This suggests that SSG6 is localized at the chloroplast envelope membrane. We generated stable transgenic rice plants expressing SSG6-GFP under the control of the maize UBIQUITIN1 promoter. In transgenic SSG6-GFP rice plants, SSG6GFP fluorescence was detected in pollen, endosperm, and pericarp (Fig. 6). In pollen, SSG6-GFP fluorescence was observed as numerous small ring-like structures (Fig. 6, A–F). Differential interference contrast images of pollen showed that the ring-like GFP fluorescence surrounded rod-shaped structures (Fig. 6, E and F); these structures are likely to be SGs, because their morphologies are consistent with the iodine-stained SGs of wild-type pollen (Fig. 3, A and C). Pollen SGs surrounded by SSG6-GFP were composed of a single starch granule (Fig. 3, A–F). This means that SGs in pollens were simple SGs. In developing endosperm and pericarp, SSG6-GFP fluorescence was detected at the membrane of amyloplasts, which contained compound SGs (Fig. 6, G–L). Taken together, these data suggest that SSG6 is localized at the amyloplast membrane in several tissues. Agronomic Traits of ssg6

To examine whether the ssg6 mutation affects rice growth under natural conditions, we grew the thirdPlant Physiol. Vol. 170, 2016

backcross generation (BC3F2) of ssg6 and wild-type plants in a paddy field and compared major agronomic traits, including duration to heading, culm length, panicle length, total panicle number, number of spikelets per panicle, number of sterile spikelets per panicle, and total panicle weight. Under natural conditions, whole-plant phenotypes did not differ significantly between BC3F2 ssg6 and the wild type (Fig. 7A). Of the seven agronomic traits measured, four traits showed statistically significant differences between BC3F2 ssg6 and the wild type (Fig. 7, B–H). The duration to heading, culm length, and number of panicles were reduced slightly in BC3F2 ssg6 compared with those in the wild type. Total panicle weight also was lower in BC3F2 ssg6 than in the wild type; this is considered to be due to the reduced number of total panicles and a slight reduction of seed weight in BC3F2 ssg6 compared with the wild type (Fig. 1D). SGs in Pollens of ssg4 ssg6 Double Mutants

ssg4 and ssg6 mutants had similar phenotypes: enlarged SGs in endosperms and spherical SGs in pollens. Table IV. Segregation of the selfed progeny of ssg4+/2ssg62/2 Selfed progeny of ssg4+/2ssg62/2 plants were genotyped for the ssg4 locus. Parental Genotype

ssg4+/2ssg62/2

No. of No. of No. of ssg42/2ssg62/2 ssg4+/2ssg62/2 ssg4+/+ssg62/2 Total

0

88

93

181 1453

Matsushima et al.

in the ssg4 mutant (Fig. 1, E–I; Matsushima et al., 2014). SGs in ssg6 pollens are sphere shaped, similar to those in the ssg4 mutant (Fig. 3, A–D; Matsushima et al., 2014). By contrast, SGs in ssg6 root cap and pericarp are similar to those of the wild type but differ from those in the ssg4 mutant (Supplemental Fig. S5; Matsushima et al., 2014). This indicates that SG size is controlled by SSG4 in all tissues, whereas the requirement for SSG6 activity in regulating SG size varies in different tissues. The ssg6 mutation introduces a premature stop codon into the coding region, which generates a truncated protein whose length is approximately 90% shorter than that of the wild type (Fig. 4B; Supplemental Fig. S9). The missing region contains part of the transmembrane domain and all of the Aminotran_1_2 domain. Therefore, the ssg6 mutation is probably a null mutation, although ssg6 phenotypes appear to be mild. The seeds of ssg4 were more chalky than those of ssg6 (Fig. 1, A and B; Supplemental Fig. S1A; Matsushima et al., 2014), and not all endosperm SGs were enlarged in ssg6 (Fig. 2, N–P). During endosperm development in ssg6, the number of smaller SGs increased and coexisted with enlarged SGs in the same cells (Fig. 2, G– L). By contrast, an increase in smaller SGs was not obvious in ssg4 developing endosperm (Matsushima et al., 2014). The smaller SGs may compensate for the starch amount in ssg6. Consistent with this, the total starch amount in ssg6 endosperm did not change during development (Fig. 1J), whereas the total starch amount in ssg4 endosperm decreased during development (Matsushima et al., 2014). The coexistence of small and large SGs in the same cells also was observed in barley and wheat cereal crops (Evers, 1973; French, 1984; Jane et al., 1994; Matsushima et al., 2010). This is called bimodal SGs. A phylogenetic analysis of Poaceae shows that only five genera contain species that develop bimodal SGs: Brachypodium, Triticum, Hordeum, Elymus, and Bromus (Tateoka, 1962; Matsushima et al., 2013). This suggests that bimodal SGs are phylogenetically specific for Poaceae. The molecular mechanisms that regulate natural variations in SG morphologies are currently unknown. The activities of SSG6 homologs in these cereal crops may be involved in this natural morphological variation. The size of the starch granule is a target for breeding and biotechnology programs to improve starch uses in industrial applications (Lindeboom et al., 2004). However, the enlargement of SGs in ssg6 does not result in the direct expansion of starch granules (Supplemental Fig. S1C). SGs in ssg6 are compound SGs that separate into starch granules during processing. Therefore, the

Table V. Segregation of the selfed progeny of ssg42/2ssg6+/2 2/2

Selfed progeny of ssg4 locus. Parental Genotype

+/2

ssg6

plants were genotyped for the ssg6

No. of No. of No. of ssg42/2ssg62/2 ssg42/2ssg6+/2 ssg42/2ssg6+/+ Total

ssg42/2ssg6+/2

0

175

164

339

We tried to construct ssg42/2ssg62/2 double mutant to investigate how the two mutations interact. In the F2 populations between ssg4 and ssg6, ssg4+/2ssg62/2 and ssg42/2ssg6+/2 mutants were obtained. The selfed progeny of ssg4+/2ssg62/2 and ssg42/2ssg6+/2 plants were genotyped at the ssg4 and ssg6 loci (Tables IV and V). Out of 181 progeny of ssg4+/2ssg62/2 and 339 progeny of ssg42/2ssg6+/2, no double homozygous mutant (ssg42/2ssg62/2) was obtained. This raised the possibility that the transmission of male and female gametophytes was affected by the double mutations. We checked the transmission efficiencies of male and female gametophytes of ssg4+/2ssg62/2 and ssg42/2ssg6+/2 by reciprocal crosses with cv Koshihikari (Tables VI and VII). In both ssg4+/2ssg62/2 and ssg42/2ssg6+/2, the transmission of ssg42ssg62 male gametophytes was significantly impaired. In contrast, the transmission of ssg42ssg62 female gametophytes was not affected. This suggested that male gametophytes of the ssg42ssg62 double mutant reduced activity compared with the ssg4 and ssg6 single mutant gametophytes. Iodine-stained pollens obtained from ssg4+/2ssg62/2 and ssg42/2ssg6+/2plants were segregated with respect to the SG morphologies (Fig. 8). Some pollens contained spherical simple SGs like ssg4 and ssg6 single mutants (Fig. 8, A and B). Other pollens showed enlarged compound SGs (Fig. 8, C and D). TEM images confirmed that the spherical SGs were composed of a single starch granule; in contrast, the enlarged SGs were assemblies of starch granules (Fig. 8, E–H). The segregation of pollens with simple and compound SGs was approximately 1:1 in both ssg4+/2ssg62/2 and ssg42/2ssg6+/2 plants (Fig. 8I).

DISCUSSION Regulation of SG Size by SSG6

In this study, we used our rapid method to observe SGs in rice endosperm and isolated the ssg6 mutant from a population of mutagenized cv Koshihikari rice. SGs in ssg6 endosperm were enlarged, similar to those

Table VI. Genetic transmission analysis of the ssg4+/2ssg62/2 mutation Parental Cross +/2

2/2

cv Koshihikari 3 ssg4 ssg6 ssg4+/2ssg62/2 3 cv Koshihikari a

1454

No. of ssg4+/+ssg6+/2

No. of ssg4+/2ssg6+/2

TEa

58 23

2 16

3.4% 69.6%

TE (transmission efficiency) = number of ssg4+/2ssg6+/2/number of ssg4+/+ssg6+/2 3 100. Plant Physiol. Vol. 170, 2016

Rice Mutant with Enlarged Starch Grains

Table VII. Genetic transmission analysis of the ssg42/2ssg6+/2 mutation Parental Cross 2/2

No. of ssg4+/2ssg6+/+

No. of ssg4+/2ssg6+/2

TEa

81 16

5 14

6.2% 87.5%

+/2

cv Koshihikari 3 ssg4 ssg6 ssg42/2ssg6+/2 3 cv Koshihikari a

TE (transmission efficiency) = number of ssg4+/2ssg6+/2/number of ssg4+/2ssg6+/+ 3 100.

ssg6 mutant is not the direct target of breeding programs. In contrast, in the case of simple SGs, the size of SGs is industrially important, because simple SGs do not separate into starch granules and the size of simple SGs is equal to that of the starch granule. Therefore, SSG6 homolog mutants of other crops that develop simple SGs, such as maize, sorghum, and barley, will achieve engineered starch granule enlargement.

Subcellular Localization of SSG6

SSG6-GFP fluorescence is detected on the envelope membrane of amyloplasts and chloroplasts (Fig. 6; Supplemental Fig. S11). This suggests that SSG6 is a membrane protein of these organelles. ACS12 also is localized at the chloroplast envelope membrane, according to the AT_CHLORO database (http:// www.grenoble.prabi.fr/at_chloro/), which collates proteomics information for intraplastidic protein localization in Arabidopsis (Ferro et al., 2010). The putative transmembrane domains of SSG6 and ACS12 should be functional. The membrane topology and subplastid localization of SSG6 protein are crucial to understand SSG6

function. Experiments to obtain such information are necessary in the future. Possible Functions of the SSG6 Protein

ACS catalyzes the conversion of S-adenosyl-Met to 1-aminocyclopropane-1-carboxylic acid, which is the precursor of ethylene. This reaction is the rate-limiting step in the ethylene biosynthetic pathway (Bleecker and Kende, 2000). ACS isozymes are well studied (Yamagami et al., 2003; Tsuchisaka et al., 2009). Arabidopsis ACS10 and ACS12 were reported to not have ACS activity; therefore, these proteins have been largely excluded from previous analyses (Yamagami et al., 2003; Tsuchisaka et al., 2009). Here, we show that SSG6 has highest homology with ACS10 and ACS12 (Fig. 4C). SSG6, ACS10, and ACS12 contain putative transmembrane domains, whereas no other ACSs contained these domains (Fig. 4D). Therefore, ACS10 and ACS12 are functional homologs of SSG6 in Arabidopsis. ACS10 and ACS12 were shown to be aminotransferases (Yamagami et al., 2003). Therefore, SSG6 also may have aminotransferase activity. OsAlaAT1 is a cytosolic aminotransferase that is responsible for the conversion of pyruvate to Ala in rice Figure 8. Segregation of pollens from ssg4+/2ssg62/2 and ssg42/2ssg6+/2 plants. A and B, Iodine-stained pollens containing spherical simple SGs from ssg4+/2ssg62/2 and ssg42/2ssg6+/2, respectively. Bars = 10 mm. C and D, Iodine-stained pollens containing enlarged compound SGs from ssg4+/2ssg62/2 and ssg42/2ssg6+/2, respectively. Bars = 10 mm. E and F, TEM images of simple SGs from ssg4+/2ssg62/2 and ssg42/2ssg6+/2, respectively. Bars = 1 mm. G and H, TEM images of compound SGs from ssg4+/2ssg62/2 and ssg42/2ssg6+/2, respectively. Bars = 1 mm. Gaps between starch granules are indicated by red arrowheads. I, Segregation of pollens with simple and compound SGs. Mature anthers from the ssg4+/2 ssg62/2 and ssg42/2 ssg6+/2 plants were disrupted with forceps. The released pollens were stained with iodine solution and examined with a microscope.

Plant Physiol. Vol. 170, 2016

1455

Matsushima et al.

endosperm (Yang et al., 2015). The osalaat1 mutation decreased the expression of starch biosynthetic genes and generated floury endosperm. Pyruvate-related metabolism was suggested to affect the substrate supply for starch synthesis (Yang et al., 2015). Further characterization of SSG6 will uncover the unknown connection between aminotransferase and starch metabolism. The rice expression profile database RiceXPro (http:// ricexpro.dna.affrc.go.jp/) indicates that SSG6 is expressed ubiquitously in nonstorage and storage organs (Sato et al., 2011). A variegated phenotype was observed in the third leaves of ssg6 mutants (Fig. 3, F and G). Chloroplasts were enlarged in third and flag leaves of ssg6 (Fig. 3, H–N; Supplemental Fig. S7). Therefore, SSG6 has a role in regulating chloroplast size in nonstorage organs. This may be related to the slight reduction in duration to heading, culm length, and number of panicles in ssg6 grown in a paddy field (Fig. 7). Neither acs10 nor acs12 mutants have been analyzed. Phenotypic analysis of those mutants will clarify the biological functions of these gene groups in vivo.

Interaction of ssg4 and ssg6 Mutations

Half of the pollens of ssg42/2ssg6+/2 and ssg4+/2ssg62/2 showed severe phenotypes with enlarged compound SGs (Fig. 8I). Because the other half of pollens contained spherical simple SGs like ssg4 and ssg6 single mutant pollens, pollens with the enlarged compound SGs will be in the ssg42ssg62 genotype. This means that ssg4 and ssg6 mutations act synergistically in the size control of pollen SGs. The ssg42ssg62 pollens should develop normally but have severely reduced activity of transmission. That can explain the segregation distortion of selfed progeny of ssg42/2ssg6+/2 and ssg4+/2ssg62/2 (Tables IV and V). However, we could not exclude the possibility that postfertilization lethality, such as embryo lethality, contributes to the unavailability of the double homozygous mutant (ssg42/2ssg62/2), because the male transmission of ssg42ssg62 was not zero (Tables VI and VII). An Arabidopsis knockout mutant of the SSG4 homolog (At2g25660) showed embryo lethality (Meinke et al., 2008). It is possible that the synergistic effect of ssg4 and ssg6 mutations may occur in other tissues in addition to the pollens. Amyloplasts and chloroplasts both develop from proplastids (Sakamoto et al., 2008). Chloroplast size is controlled by the chloroplast binary fission division machinery, especially by the ring structures at the division sites (Miyagishima, 2011). Rice mutants defective in the ring structures develop enlarged and dumbbellshaped chloroplasts in leaves, but they do not develop enlarged amyloplasts (Yun and Kawagoe, 2009; Kamau et al., 2015). Therefore, ssg4 and ssg6 mutants are unique in that they affect both chloroplasts and amyloplasts. The mechanism to develop compound SGs in ssg4 2 ssg6 2 pollens was unknown. Proplastids before starch accumulation may be already enlarged in 1456

ssg42ssg62 pollens and retain multiple priming sites of starch granule synthesis. We believe that ssg4 and ssg6 mutants are useful for future starch breeding programs and for basic understanding of the mechanisms regulating plastid size in plants.

MATERIALS AND METHODS Plant Materials and Growth Conditions Rice (Oryza sativa) ssp. japonica ‘Koshihikari’ and ‘Nipponbare’ and ssp. indica ‘Kasalath’ were used as wild-type plants. The ssg6 mutant was isolated from a mutagenized cv Koshihikari population. Mutagenization was carried out by soaking seeds in 1.5% (v/v) methanesulfonic acid ethyl ester (Nacalai Tesque; 15519-54). The M2 lines derived from a single M1 plant were grown, and M2 seeds were collected from individual M1 plants after self-fertilization. We backcrossed the ssg6 mutant with the wild type, and their progeny with ssg6 phenotypes were used in this study. The ssg6 mutant was also backcrossed with cv Nipponbare for transformation of complementation test. Rice plants were grown at an experimental paddy field of the Institute of Plant Science and Resources, Okayama University, under natural conditions. Transgenic plants were grown at 28°C in a greenhouse.

Isolation of ssg6 by the Rapid Observation Method Screening was carried out with at least five seeds from each M2 line. Endosperm thin sections from M2 seeds were prepared, and SGs were observed by the previously developed method (Matsushima et al., 2010).

Characterization of Grain Appearance, Sizes, and Starch Amounts Matured dry seeds and maturing young seeds were photographed with a macromicroscope (MVX10; Olympus) and a digital camera (DP72; Olympus). The sizes of grains were measured with a Vernier caliper. The total amount of starch was measured by enzymatic methods using the Total Starch Assay Kit (Megazyme International).

Thermal Properties of Starch Thermal properties of starch were analyzed using a differential scanning calorimeter (DSC-6100; Seiko Instruments). The procedures were the same as in our previous study (Matsushima et al., 2014).

Chain-Length Distribution of Amylopectin The chain-length distribution of amylopectin was analyzed using fluorescence labeling and capillary electrophoresis (Beckman Coulter). The procedures were the same as in our previous study (Matsushima et al., 2010).

Semiquantification of Sugars in Mature Seeds After separating husks from seeds, the brown rice (50 seeds) of cv Koshihikari and the ssg6 mutant were bulked separately and then crushed using a Retsch mixer mill (MM301) at a frequency of 20 Hz21 for 2 min at 4°C. We prepared five pooled samples for each genotype. Extracts from the bulked seeds (equivalent to 20 mg) of each sample were subjected to gas chromatography-time-of-flightmass spectrometry as described previously (Kusano et al., 2007). Peaks of each sugar in each analyte were identified by comparing retention indices and the mass spectra of the corresponding authentic standards.

Observation of SGs and Chloroplasts by Thin-Section Microscopy with Technovit 7100 Resin For matured dry seeds, approximately 1-mm cubic blocks were cut from the center region of the endosperm and fixed in a solution containing 5% (v/v) formalin, 5% (v/v) acetic acid, and 50% (v/v) ethanol for at least 12 h at room Plant Physiol. Vol. 170, 2016

Rice Mutant with Enlarged Starch Grains

temperature. For maturing endosperm, approximately 1-mm3 blocks were cut out from the maturing endosperm at 3, 5, and 7 DAF and fixed in 3% (v/v) glutaraldehyde in 20 mM cacodylate buffer (pH 7.4) for at least 24 h at 4°C. For root caps, the seminal root tips (1 mm) were cut out and fixed in the same buffer as for the maturing endosperm. To observe chloroplasts, the middle region of leaves was sampled. The procedures of resin embedding using Technovit 7100 resin (Kulzer) were described previously (Matsushima et al., 2010). Thin sections (1 mm) were prepared using an ultramicrotome (EM UC7; Leica Microsystems) and diamond knives. Staining procedures were described previously (Matsushima et al., 2010). Quantification of images was performed with ImageJ 1.46r software (http://rsbweb.nih.gov/ij/).

Observation of Pollen Grains To obtain SGs in pollen grains, anthers just before anthesis were disrupted with forceps in the diluted Lugol solution (iodine/potassium iodide solution; MP Biomedicals) on a glass slide. Observation was done using a microscope (AX70; Olympus).

TEM Observation The middle region of leaves and 3-DAF endosperms were fixed in the fix buffer consisting of 4% (w/v) paraformaldehyde and 2% (v/v) glutaraldehyde in 50 mM cacodylate buffer (pH 7.4) at 4°C overnight followed by postfixation with 2% osmium tetroxide at 4°C for 3 h. In the case of pollens, mature anthers were disrupted with forceps in the fix buffer on a glass slide. The released pollens were suspended in the fix buffer followed by postfixation. The procedures to obtain the ultra-thin sections (70 nm) were described previously (Matsushima et al., 2014). Observation was carried out using a transmission electron microscope (JEM-1400Plus; JEOL) at an acceleration voltage of 80 kV. Digital images were taken with a CCD camera (VELETA; Olympus Soft Imaging Solutions).

Map-Based Cloning and Microarray Analysis to Determine the ssg6 Locus To map the SSG6 gene, we constructed an F2 population derived from a cross between the ssg6 mutant and cv Kasalath. To select ssg6 mutant seeds from the F2 population, endosperm thin sections of each F2 seed were examined by the rapid method that was developed previously (Matsushima et al., 2010, 2014). The genomic DNA of these ssg6 mutants was individually isolated and analyzed using simple sequence length polymorphism markers to determine the molecular markers linked to the ssg6 phenotype (Temnykh et al., 2000; McCouch et al., 2002). Primers used for molecular markers are as follows: RM588, 59-GTTGCTCTGCCTCACTCTTG-39 and 59-AACGAGCCAACGAAGCAG-39; and RM190, 59-CTTTGTCTATCTCAAGACAC-39 and 59-TTGCAGATGTTCTTCCTGATG-39. The SSG6 locus was mapped in the 152-kb region between RM588 and RM190 on chromosome 6. Next, total RNA was prepared from 4-DAF developing seeds of cv Koshihikari and ssg6 using the RNeasy Plant Mini Kit (Qiagen). Microarray analysis was performed according to Agilent Oligo DNA Microarray Hybridization protocols using the Rice Oligo DNA Microarray 44K RAP database (G2519F; Agilent Technologies) with four biological replicates. The hybridized slides were scanned using a DNA microarray scanner (G2565CA; Agilent Technologies). Signal intensities were extracted by Feature Extraction software (Agilent Technologies). The signal intensities of the Os06g0130400 gene, which is located in the 152-kb region, were reduced significantly in ssg6 compared with the wild type. We determined the nucleotide sequence of the Os06g0130400 gene and identified a single-base substitution.

RACE Experiment The RACE experiment was performed using the SMARTer RACE cDNA amplification kit (Clontech) according to the manufacturer’s instructions. The first-strand cDNA was synthesized from total RNA of 4-DAF developing seeds of cv Koshihikari. To determine the 59 untranslated region of the Os06g0130400 gene, the following gene-specific primer was used: 59-CGGTTGCGCTTGATGTAGAGCT-39. To determine the 39 untranslated region, the following gene-specific primer was used: 59-GAGAGATTGCGCAAGATGTACC-39. The sequence of the obtained PCR fragment was used as a template for direct sequencing using the Big Dye Terminator version 3.1 Cycle Sequencing Kit (Applied Biosystems) and a genetic analyzer (Applied Biosystems 3130xl). Plant Physiol. Vol. 170, 2016

Plasmid Construction Because the regeneration efficiency of cv Koshihikari is lower than that of cv Nipponbare, the ssg6 mutant was backcrossed with cv Nipponbare to change the genetic background. For complementation, the Os01g0179400 cDNA fragment was amplified from the full-length cDNA clone (AK065212; National Institute of Agrobiological Sciences DNA Bank; http://www.dna.affrc.go.jp/) using the following primers: 59-TCAGTCGACTGGATCATGAGGCGGAGCGGCAACGGCGGCG-39 and 59-GTGCGGCCGCGAATTTCAATGATTCGGTTTATGGCTATCT-39. The fragment was cloned into the BamHI and EcoRI sites of the pENTR2B entry vector (Invitrogen) using the In-Fusion Cloning Kit (Clontech). Unfortunately, the resulting plasmid (pENTRSSG6cDNA-defective) lacks the GC-rich region of the first half of the Os01g0179400 cDNA. Therefore, the missing region was amplified again from AK065212 using primers 59-AACCAATTCAGTCGAATGAGGCGGAGCGGCAACGGCGGCG-39 and 59-GTAGGCGGCGAGGCCTCTGATGGAG-39. The amplified fragment was inserted into the SalI and SacI sites of pENTRSSG6cDNA-defective. The SalI site is located in the vector-derived region, and the SacI site is in the middle of the cDNA. The resulting plasmid (pENTRSSG6cDNA-full) is used for the LR recombination reaction with the destination vector pIPKb002 (Himmelbach et al., 2007) using the Gateway system (Invitrogen). The resulting plasmid (pK02SSG6cDNA-full) was then introduced into the ssg6 mutant in the cv Nipponbare background using an Agrobacterium tumefaciens-mediated method (Hiei et al., 1994). To construct transgenic plants expressing the SSG6 protein fused with GFP, SSG6 cDNA was fused with the GFP gene. SSG6 cDNA without a stop codon was amplified using AK065212 as a template and the following primers: 59TCAGTCGACTGGATCTGGACTCGTTTGCCGCCAGCAACAT-39 and 59GTGCGGCCGCGAATTATGATTCGGTTTATGGCTATCTGTA-39. The amplified fragment was inserted into the BamHI and EcoRI sites of the pENTR2B entry vector. The resulting plasmid is used for the LR recombination reaction with the destination vector pGWB405 (Nakagawa et al., 2007). In the resulting plasmid (pGWBSSG6-GFP), SSG6 cDNA is followed by the GFP gene. The SSG6-GFP gene was amplified from pGWBSSG6-GFP as a template using the following primers: 59-TCAGTCGACTGGATCATGAGGCGGAGCGGCAACGGCGGCG-39 and 59-GTGCGGCCGCGAATTTTACTTGTACAGCTCGTCCATGC-39. The fragment was cloned into the BamHI and EcoRI sites of the pENTR2B entry vector. The resulting plasmid was named pENTRSSG6-GFP. pENTRSSG6GFP was used for the LR reaction with the destination vector pIPKb002. The resulting plasmid (pK02SSG6-GFP) was then introduced into cv Nipponbare. pENTRSSG6-GFP was also used for the LR reaction with the destination vector pGWB602 (Nakamura et al., 2010) to introduce the SSG6-GFP gene under the control of the cauliflower mosaic virus 35S promoter. The resulting plasmid (p602SSG6-GFP) was used for the A. tumefaciens-mediated transient transformation of Nicotiana benthamiana.

A. tumefaciens-Mediated Transient Transformation of N. benthamiana A. tumefaciens-mediated transient transformation was carried out as described previously (Matsushima et al., 2014).

Phylogenetic Analysis SSG6 homologous sequences were searched through BLAST in the TAIR and RAP databases. Sequences were aligned with ClustalW (http://www.genome. jp/tools/clustalw/). Trees were constructed on conserved positions of the alignment by clustered protein sequences from plants with the neighbor-joining algorithm as implemented in MEGA 5.2 with pairwise deletion for gap filling (Tamura et al., 2011). To test inferred phylogeny, we used bootstraps with 1,000 bootstrap replicates.

Analysis of Amylose Content The polished rice grains were pulverized by Cyclotec 1093 (FOSS) to obtain powder. The powder (100 mg) was suspended in 5 mL of 50 mM NaOH and subjected to AutoAnalyzer (Technicon). In the AutoAnalyzer, the sample was mixed with 400 mM NaOH and heated at 93°C for gelatinization. The gelatinized sample was mixed with 30 mM citric acid and 85 mM acetic acid buffer and stained with iodine solution (3 mM potassium iodide and 0.2 mM iodine). Finally, the optical density at 620 nm was measured. A standard curve was made using potato (Solanum tuberosum) amylose standard (Sigma; A0512). 1457

Matsushima et al.

Detection of GFP Signals in Endosperms and Pericarps of SSG6-GFP Transgenic Plants Developing seeds at 3 DAF without husks were embedded in 5% (w/v) agarose and cross sectioned through the middle portion of the seed in 120-mmthick sections with a vibrating blade microtome (VT-1200S; Leica Microsystems). The sections were examined using the laser scanning confocal microscope (FV1000; Olympus).

Measurement of Agronomic Traits The third backcross generation of ssg6 (BC3F2) was grown in an experimental field at the Institute of Plant Science and Resources, Okayama University. In 2014, germinated seeds were sown on April 10, and the seedlings were transplanted into the paddy field about 30 d later with a distance of 18 cm between plants and 30 cm between rows. The plants were raised according to standard practices for the time of year in Japan. Phenotypic traits were measured after harvest except for the duration to heading. The seven phenotypic traits were duration to heading, culm length, panicle length, number of panicles, number of spikelets per panicle, number of sterile spikelets per panicle, and total panicle weight.

Supplemental Figure S2. Semiquantification of sugars in wild-type cv Koshihikari and the ssg6 mutant. Supplemental Figure S3. TEM images of SGs at 3 DAF. Supplemental Figure S4. Accumulation of starch biosynthetic enzymes in developing seeds of wild-type cv Koshihikari and ssg6 mutants. Supplemental Figure S5. SG morphologies in ssg6 root cap and pericarp. Supplemental Figure S6. Number of chloroplasts per cell and number of starch granules per chloroplast of third and flag leaves. Supplemental Figure S7. Chloroplast morphologies in flag leaves of ssg6. Supplemental Figure S8. Microarray analysis of gene expression in wildtype cv Koshihikari and ssg6. Supplemental Figure S9. SSG6 coding DNA sequence and its translated amino acid sequence in wild-type cv Koshihikari. Supplemental Figure S10. Multiple sequence alignment of SSG6 homologs from Arabidopsis and rice. Supplemental Figure S11. Chloroplast membrane localization of SSG6GFP.

Detection of ssg4 and ssg6 Mutations The ssg4 mutation sites were detected by the derived cleaved-amplified polymorphic sequence primers designed as follows: 59-ACCTAAGGGTGTCTTAACTTTTGAAACT-39 and 59-AAAACCACAGTAACCATGAGAATGT-39. The PCR conditions were as follows: 94°C for 2 min and 35 cycles of 94°C for 30 s, 52°C for 45 s, and 68°C for 1 min. The PCR product was digested with SpeI, and PCR products were subsequently separated by 15% PAGE and detected with ethidium bromide staining. In the case of ssg4, a PCR product (140 bp) was digested into 114 and 26 bp. In the case of the wild type, the PCR product was not digested. To detect the ssg6 mutation, the derived cleaved-amplified polymorphic sequence primers were designed as follows: 59-CTCATCCCCTGCGCCCTCTTCTACTTCTGC-39 and 59-GCGTAGTAGAGCGAGTCGTCGTC-39. The PCR conditions were as follows: 95°C for 2 min and 35 cycles of 95°C for 30 s, 57°C for 45 s, and 72°C for 1 min. The PCR product was digested with NheI, and PCR products were subsequently separated by 15% PAGE and detected with ethidium bromide staining. In the case of ssg6, a PCR product (218 bp) was digested into 189 and 29 bp. In the case of the wild type, the PCR product was not digested.

Total Protein Extraction and Immunoblotting Two seeds at 11 DAF from wild-type cv Koshihikari and ssg6 mutants were ground in 93 volume of urea-SDS sample buffer consisting of 0.125 M Tris-HCl, pH 6.8, 8 M urea, 4% (w/v) SDS, 5% (v/v) b-mercaptoethanol, and 0.02% (w/v) Bromophenol Blue supplemented with 10 mL mL21 protease inhibitor cocktail (Sigma). Samples were extracted for 1 h at room temperature on a rotating platform and centrifuged at 20,000g at room temperature for 10 min to remove starch. The supernatants (7.5 mL) were subjected to 7.5% SDS-PAGE. After SDSPAGE, the proteins were transferred electrophoretically to a polyvinylidene difluoride membrane (Millipore) by transblotter (Nihon Eido). The rest of the western-blotting procedure was performed as described by Crofts et al. (2012). Primary antibodies were used at the following dilutions: anti-SSI (Fujita et al., 2006) at 1:2,000, anti-SSIIa at 1:1,000, anti-SSIIIa (Crofts et al., 2012) at 1:2,000, anti-SSIVb at 1:1,000, anti-GBSSI (Fujita et al., 2006) at 1:5,000, anti-BEI (Nakamura et al., 1992) at 1:2,000, anti-BEIIa at 1:3,000, anti-BEIIb (Nakamura et al., 1992) at 1:3,000, anti-PUL (Nakamura et al., 1996) at 1:1,000, and anti-PHO1 (Satoh et al., 2008) at 1:1,000. The accession number of the SSG6 full-length cDNA sequence is LC050636 in the GenBank/EMBL/DNA Data Bank of Japan databases. All other sequence data used in this article can be found in TAIR and RAP databases.

Supplemental Data The following supplemental materials are available. Supplemental Figure S1. Polished grain appearance, number and size of starch granules in SGs, amylopectin chain-length distribution, and amylose content of ssg6 mutants. 1458

ACKNOWLEDGMENTS We thank Makoto Kobayashi (RIKEN) and Kazuki Saito (RIKEN) for metabolite profiling analysis, Dr. Tsuyoshi Nakagawa (Shimane University) for providing Gateway vectors (pGWB405 and pGWB602) containing the bar gene identified by Meiji Seika Kaisha, the Leibniz Institute of Plant Genetics and Crop Plant Research for providing another Gateway vector (pIPKb002), and the Biotechnology Center of Akita Prefectural University for technical assistance. Received December 1, 2015; accepted January 18, 2016; published January 20, 2016.

LITERATURE CITED Ai Y, Medic J, Jiang H, Wang D, Jane JL (2011) Starch characterization and ethanol production of sorghum. J Agric Food Chem 59: 7385–7392 Bleecker AB, Kende H (2000) Ethylene: a gaseous signal molecule in plants. Annu Rev Cell Dev Biol 16: 1–18 Buléon A, Colonna P, Planchot V, Ball S (1998) Starch granules: structure and biosynthesis. Int J Biol Macromol 23: 85–112 Bustos R, Fahy B, Hylton CM, Seale R, Nebane NM, Edwards A, Martin C, Smith AM (2004) Starch granule initiation is controlled by a heteromultimeric isoamylase in potato tubers. Proc Natl Acad Sci USA 101: 2215–2220 Crofts N, Abe K, Aihara S, Itoh R, Nakamura Y, Itoh K, Fujita N (2012) Lack of starch synthase IIIa and high expression of granule-bound starch synthase I synergistically increase the apparent amylose content in rice endosperm. Plant Sci 193-194: 62–69 Crumpton-Taylor M, Grandison S, Png KM, Bushby AJ, Smith AM (2012) Control of starch granule numbers in Arabidopsis chloroplasts. Plant Physiol 158: 905–916 Evers AD (1973) The size distribution among starch granules in wheat endosperm. Starke 25: 303–304 Ferro M, Brugière S, Salvi D, Seigneurin-Berny D, Court M, Moyet L, Ramus C, Miras S, Mellal M, Le Gall S, et al (2010) AT_CHLORO, a comprehensive chloroplast proteome database with subplastidial localization and curated information on envelope proteins. Mol Cell Proteomics 9: 1063–1084 French D (1984) Organization of starch granules. In Starch Chemistry and Technology, Ed 2. Academic Press, New York, pp 183–242 Fujita N (2014) Starch biosynthesis in rice endosperm. AGri-Bioscience Monographs 4: 1–18 Fujita N, Yoshida M, Asakura N, Ohdan T, Miyao A, Hirochika H, Nakamura Y (2006) Function and characterization of starch synthase I using mutants in rice. Plant Physiol 140: 1070–1084 Gutiérrez OA, Campbell MR, Glover DV (2002) Starch particle volume in single- and double-mutant maize endosperm genotypes involving the soft starch (h) gene. Crop Sci 42: 355–359 Plant Physiol. Vol. 170, 2016

Rice Mutant with Enlarged Starch Grains

Hancock RD, Tarbet BJ (2000) The other double helix: the fascinating chemistry of starch. J Chem Educ 77: 988–992 Hiei Y, Ohta S, Komari T, Kumashiro T (1994) Efficient transformation of rice (Oryza sativa L.) mediated by Agrobacterium and sequence analysis of the boundaries of the T-DNA. Plant J 6: 271–282 Himmelbach A, Zierold U, Hensel G, Riechen J, Douchkov D, Schweizer P, Kumlehn J (2007) A set of modular binary vectors for transformation of cereals. Plant Physiol 145: 1192–1200 Iwai T, Miyasaka A, Seo S, Ohashi Y (2006) Contribution of ethylene biosynthesis for resistance to blast fungus infection in young rice plants. Plant Physiol 142: 1202–1215 Jane JL, Kasemsuwan T, Leas S, Ames IA, Zobel H, Darien IL, Robyt JF, Ames IA (1994) Anthology of starch granule morphology by scanning electron microscopy. Starke 46: 121–129 Ji Q, Oomen RJ, Vincken JP, Bolam DN, Gilbert HJ, Suurs LC, Visser RG (2004) Reduction of starch granule size by expression of an engineered tandem starch-binding domain in potato plants. Plant Biotechnol J 2: 251–260 Kamau PK, Sano S, Takami T, Matsushima R, Maekawa M, Sakamoto W (2015) A mutation in GIANT CHLOROPLAST encoding a PARC6 homolog affects spikelet fertility in rice. Plant Cell Physiol 56: 977–991 Kusano M, Fukushima A, Kobayashi M, Hayashi N, Jonsson P, Moritz T, Ebana K, Saito K (2007) Application of a metabolomic method combining one-dimensional and two-dimensional gas chromatographytime-of-flight/mass spectrometry to metabolic phenotyping of natural variants in rice. J Chromatogr B Analyt Technol Biomed Life Sci 855: 71–79 Lindeboom N, Chang PR, Tyler RT (2004) Analytical, biochemical and physicochemical aspects of starch granule size, with emphasis on small granule starches: a review. Starke 56: 89–99 Malinski E, Daniel JR, Zhang XX, Whistler RL (2003) Isolation of small starch granules and determination of their fat mimic characteristics. Cereal Chem 80: 1–4 Matsushima R, Maekawa M, Fujita N, Sakamoto W (2010) A rapid, direct observation method to isolate mutants with defects in starch grain morphology in rice. Plant Cell Physiol 51: 728–741 Matsushima R, Maekawa M, Kusano M, Kondo H, Fujita N, Kawagoe Y, Sakamoto W (2014) Amyloplast-localized SUBSTANDARD STARCH GRAIN4 protein influences the size of starch grains in rice endosperm. Plant Physiol 164: 623–636 Matsushima R, Yamashita J, Kariyama S, Enomoto T, Sakamoto W (2013) A phylogenetic re-evaluation of morphological variations of starch grains among Poaceae species. J Appl Glycosci 60: 37–44 McCouch SR, Teytelman L, Xu Y, Lobos KB, Clare K, Walton M, Fu B, Maghirang R, Li Z, Xing Y, et al (2002) Development and mapping of 2240 new SSR markers for rice (Oryza sativa L.). DNA Res 9: 199–207 Meinke D, Muralla R, Sweeney C, Dickerman A (2008) Identifying essential genes in Arabidopsis thaliana. Trends Plant Sci 13: 483–491 Miyagishima SY (2011) Mechanism of plastid division: from a bacterium to an organelle. Plant Physiol 155: 1533–1544 Nakagawa T, Suzuki T, Murata S, Nakamura S, Hino T, Maeo K, Tabata R, Kawai T, Tanaka K, Niwa Y, et al (2007) Improved Gateway binary vectors:

Plant Physiol. Vol. 170, 2016

high-performance vectors for creation of fusion constructs in transgenic analysis of plants. Biosci Biotechnol Biochem 71: 2095–2100 Nakamura S, Mano S, Tanaka Y, Ohnishi M, Nakamori C, Araki M, Niwa T, Nishimura M, Kaminaka H, Nakagawa T, et al (2010) Gateway binary vectors with the bialaphos resistance gene, bar, as a selection marker for plant transformation. Biosci Biotechnol Biochem 74: 1315–1319 Nakamura Y, Takeichi T, Kawaguchi K, Yamanouchi H (1992) Purification of two forms of starch branching enzyme (Q-enzyme) from developing rice endosperm. Physiol Plant 84: 329–335 Nakamura Y, Umemoto T, Ogata N, Kuboki Y, Yano M, Sasaki T (1996) Starch debranching enzyme (R-enzyme or pullulanase) from developing rice endosperm: purification, cDNA and chromosomal localization of the gene. Planta 199: 209–218 Sakamoto W, Miyagishima SY, Jarvis P (2008) Chloroplast biogenesis: control of plastid development, protein import, division and inheritance. The Arabidopsis Book 6: e0110. doi: 10.1199/tab.0110 Sato Y, Antonio BA, Namiki N, Takehisa H, Minami H, Kamatsuki K, Sugimoto K, Shimizu Y, Hirochika H, Nagamura Y (2011) RiceXPro: a platform for monitoring gene expression in japonica rice grown under natural field conditions. Nucleic Acids Res 39: D1141–D1148 Satoh H, Shibahara K, Tokunaga T, Nishi A, Tasaki M, Hwang SK, Okita TW, Kaneko N, Fujita N, Yoshida M, et al (2008) Mutation of the plastidial a-glucan phosphorylase gene in rice affects the synthesis and structure of starch in the endosperm. Plant Cell 20: 1833–1849 Stahl Y, Coates S, Bryce JH, Morris PC (2004) Antisense downregulation of the barley limit dextrinase inhibitor modulates starch granule size distribution, starch composition and amylopectin structure. Plant J 39: 599–611 Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S (2011) MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28: 2731–2739 Tateoka T (1962) Starch grains of endosperm in grass systematics. Bot Mag Tokyo 75: 377–383 Temnykh S, Park WD, Ayres N, Cartinhour S, Hauck N, Lipovich L, Cho YG, Ishii T, McCouch SR (2000) Mapping and genome organization of microsatellite sequences in rice (Oryza sativa L.). Theor Appl Genet 100: 697–712 Tsuchisaka A, Yu G, Jin H, Alonso JM, Ecker JR, Zhang X, Gao S, Theologis A (2009) A combinatorial interplay among the 1-aminocyclopropane-1carboxylate isoforms regulates ethylene biosynthesis in Arabidopsis thaliana. Genetics 183: 979–1003 Yamagami T, Tsuchisaka A, Yamada K, Haddon WF, Harden LA, Theologis A (2003) Biochemical diversity among the 1-amino-cyclopropane-1-carboxylate synthase isozymes encoded by the Arabidopsis gene family. J Biol Chem 278: 49102–49112 Yang J, Kim SR, Lee SK, Choi H, Jeon JS, An G (2015) Alanine aminotransferase 1 (OsAlaAT1) plays an essential role in the regulation of starch storage in rice endosperm. Plant Sci 240: 79–89 Yun MS, Kawagoe Y (2009) Amyloplast division progresses simultaneously at multiple sites in the endosperm of rice. Plant Cell Physiol 50: 1617–1626

1459