Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A Rapid, Direct Observation Method to Isolate Mutants with Defects in Starch Grain Morphology in Rice

A Rapid, Direct Observation Method to Isolate Mutants with Defects in Starch Grain Morphology in... Abstract Starch forms transparent grains, referred to as starch grains (SGs), in amyloplasts. Despite the simple glucose polymer composition of starch, SGs exhibit different morphologies depending on plant species, especially in the endosperm of the Poaceae family. This study reports a novel method for preparing thin sections of endosperm without chemical fixation or resin embedding that allowed us to visualize subcellular SGs clearly. Using this method, we observed the SG morphologies of >5,000 mutagenized rice seeds and were able to isolate mutants in which SGs were morphologically altered. In five mutants, named ssg (substandard starch grain), increased numbers of small SGs (ssg1–ssg3), enlarged SGs (ssg4) and abnormal interior structures of SGs (ssg5) were observed. Amylopectin chain length distribution analysis and identification of the mutated gene suggested a possible allelic relationship between ssg1, ssg2, ssg3 and the previously isolated amylose-extender (ae) mutants, while ssg4 and ssg5 seemed to be novel mutants. Compared with conventional observation methods, the methods developed here are more effective for obtaining fine images of subcellular SGs and are suitable for the observation of a large number of samples. Introduction Starch is a biologically and commercially important polymer of glucose and is synthesized to form starch grains (SGs) inside plastids (amyloplasts) of higher plants (Buléon et al. 1998, Hancock and Tarbet 2000). Starch is water insoluble and hence osmotically inactive, making it a suitable long-term storage carbohydrate for seeds and tubes of many plant species. Starch contains two major components, amylose and amylopectin. Amylose is a linear molecule consisting of α-1,4-linked d-glucose chains, whereas amylopectin contains α-1,6-linked branches. Amylopectin is a major component of starch, accounting for 65–85% of its weight (w/w) (Nakamura 2002, James et al. 2003). Despite the simple composition of glucose polymers, SGs exhibit various morphologies depending on the plant species (Harz 1880, Tateoka 1954, Tateoka 1955, Tateoka 1962, Czaja 1978, Jane et al. 1994, James et al. 2003, Shapter et al. 2008). Although SG morphological diversity was reported >100 years ago, the underlying molecular mechanisms that account for the differences in morphology have not yet been determined. Morphologically, SGs are classified as either compound or simple grains (Harz 1880). Compound grains consist of many starch granules, and simple grains are composed of a single starch granule in an amyloplast. According to previous classification studies, Poaceae species are divided into four types based on the SG morphologies of their endosperm (Tateoka 1962). Compound grain types such as rice (Oryza sativa) possess only compound grains in their endosperm. Simple grain types are further divided into two subtypes referred to as bimodal and uniform. Bimodal types produce small and much larger simple SGs that co-exist in the same cells. Uniform types possess similar sized hexagonal, pentagonal or round, simple SGs. Barley (Hordeum vulgare) and wheat (Triticum aestivum) make bimodal SGs, and uniform SGs are found in maize (Zea mays). In addition to the above three types, a number of species, such as some belonging to the Andropogoneae tribe, show the co-existence of compound and simple grains inside the same cell (Tateoka 1955). Several starch-related mutants have been isolated in cereals by forward genetic approaches based on (i) the physical characteristics of seeds; (ii) the activities of starch biosynthesis enzymes; and (iii) the chemical properties of starch (Dvonch et al. 1951, Walker and Merritt 1969, Jarvi and Eslick 1975, Satoh and Omura 1981, Satoh et al. 2003a, Satoh et al. 2003b, Kang et al. 2005, Satoh et al. 2008). In addition, mutants defective in starch biosynthesis enzymes were isolated by reverse genetic approaches using rice Tos17 tagging systems (Fujita et al. 2007, Fujita et al. 2009). Some starch-related mutants exhibit SG morphologies different from those of wild-type strains (Dvonch et al. 1951, Yano et al. 1985, Burton et al. 2002, Fujita et al. 2003, Li et al. 2007); however, a molecular explanation for these morphological differences of SGs has not been proposed. Scanning electron microscopy (SEM) is normally used to observe SGs in extracted conditions where compound SGs are broken and the interior starch granules are detached from each other (Jane et al. 1994). To acquire images of subcellular SGs, cell walls need to be removed. Therefore, cells whose cell walls collapse during slicing of the endosperm make good subjects for SEM analysis (Shapter et al. 2008). Compared with SEM, light microscopic observation of endosperm thin sections is suitable for observing subcellular SGs, because they can be observed in most cells in sections. However, chemical fixation and resin embedding processes are needed for thin sectioning, which makes it unsuitable for large sample numbers, such as those found in a genetic screen. In this study, we describe a rapid method to prepare thin sections of cereal endosperm without chemical fixation or resin embedding. Using this method, we performed a genetic screen and isolated five rice mutants defective in SG morphologies. Three of the five mutants seemed to be allelic to the previously isolated amylose-extender (ae) mutants (Yano et al. 1985, Mizuno et al. 1993) and the other two appear to be novel mutants. Compared with conventional observation methods, the methods presented here are more effective for obtaining clear images of subcellular SGs and are highly suitable for the examination of a large number of samples. Results A rapid method to prepare thin sections of endosperm We developed a rapid method to prepare thin sections of mature endosperm (Fig. 1). A mature rice seed was fitted into a truncated 200 μl pipet tip (Fig. 1A). For rice and barley, truncated 200 μl tips were appropriate. For maize, a truncated 1 ml pipet tip was suitable. The seed-embedded tip was fixed on a block trimmer that was originally developed for resin block trimming for ultramicrotomy (Fig. 1B, C). The fixed seeds on the block trimmer were manipulated under a stereo microscope (Fig. 1D). The block trimmer was held by the third and fourth fingers of the non-dominant hand (Fig. 1E), while the index finger and thumb of the dominant hand held a razor blade attached to the seed. During trimming, the blade was kept horizontal to the seed (Fig. 1E, inset). The blade was also supported by the index finger of the non-dominant hand to adjust trimming pressure. Trimming generated a smooth surface on the top of the seed, which was exposed approximately 1 mm out of the pipet tip. Thin sectioning was performed in the same way as trimming with the same hand positions, but we kept the angle of the blade at approximately 30° (Fig. 1F). The seeds were easily trimmed using a razor blade (Fig. 1G, H), and thin sections were shaved off the endosperm (Fig. 1I, J). Forceps were used to take and place thin sections onto glass slides for staining (Fig. 1K). Fig. 1 View largeDownload slide The rapid method for preparation of thin sections of rice endosperm. (A) A rice seed was inserted into a truncated, disposable pipet tip (right). The inset is the magnified image of the captured seed. Bar = 5 mm. (B, C) The seed-captured tip was fixed on a block trimmer that was originally developed for resin block trimming for ultramicrotomy. (D) The block trimmer (indicated by a red arrow) can be manipulated under a stereo microscope. (E) Positions of the hands, fingers, a razor blade and the block trimmer equipped with the seed-captured tip for trimming. The inset is the magnified side view image of the captured seed and razor blade (indicated by an arrowhead). (F) A side view of the hand positions for thin sectioning. The angle of the blade is approximately 30° (inset, arrowhead). (G–K) Seed images under the stereo microscope. The razor blade and forceps are indicated by asterisks and a white arrow, respectively. Bars = 1 mm. (G) Before trimming. (H) The smooth surface after trimming. (I–J) A thin section shaved off the endosperm using the razor blade. (K) Forceps were used to take thin sections. Fig. 1 View largeDownload slide The rapid method for preparation of thin sections of rice endosperm. (A) A rice seed was inserted into a truncated, disposable pipet tip (right). The inset is the magnified image of the captured seed. Bar = 5 mm. (B, C) The seed-captured tip was fixed on a block trimmer that was originally developed for resin block trimming for ultramicrotomy. (D) The block trimmer (indicated by a red arrow) can be manipulated under a stereo microscope. (E) Positions of the hands, fingers, a razor blade and the block trimmer equipped with the seed-captured tip for trimming. The inset is the magnified side view image of the captured seed and razor blade (indicated by an arrowhead). (F) A side view of the hand positions for thin sectioning. The angle of the blade is approximately 30° (inset, arrowhead). (G–K) Seed images under the stereo microscope. The razor blade and forceps are indicated by asterisks and a white arrow, respectively. Bars = 1 mm. (G) Before trimming. (H) The smooth surface after trimming. (I–J) A thin section shaved off the endosperm using the razor blade. (K) Forceps were used to take thin sections. When the seed-embedded tip was attached to an ultramicrotome holder, thin sections could also be obtained using a glass knife and an ultramicrotome. Thin sections prepared by hand did not have a uniform thickness; however, the thinnest part of the thin sections prepared by hand was estimated to be approximately 1 μm thick, because 1 μm thin sections cut by the ultramicrotome produced a similar light transmission to the thinnest part of the hand-prepared thin sections (data not shown). To test the effectiveness of this method, we tried to prepare thin sections of three cereal species: rice, maize and barley (Fig. 2A–C). Thin sections prepared using the rapid method were stained with iodine and examined under a microscope. The morphologies of SGs were clearly observed (Fig. 2D–F). Rice SGs showed compound grains in which smaller granules were assembled (Fig. 2D, G). Occasionally, simple grains were also observed. Maize had simple grains that were round and uniform in size (Fig. 2E, H). In barley, we observed small (approximately 5 μm in diameter) and larger (approximately 20 μm in diameter) simple grains co-existing in the same cell (Fig. 2F, I). These SG morphologies were consistent with previous reports (Tateoka 1962) and SEM images (Fig. 2J–L). Fig. 2 View largeDownload slide Starch grain (SG) morphologies of three cereal species. (A–C) Seed images of rice (Oryza sativa), maize (Zea mays) and barley (Hordeum vulgare). Bars = 1 mm. (D–F) Iodine-stained sections of endosperm cells prepared using the rapid method. Bars = 20 μm. Boxes indicated by white dotted lines are illustrated areas in G–I. (G–I) Typical morphologies of SGs are illustrated. The SGs of rice are compound grains composed of many smaller granules. Maize SGs are the uniform type that are similar in size and round shaped. Barley SG is a bimodal type containing small and larger grains co-existing in the same cell. Red lines indicate cell walls. (J–L) SEM images. Bars = 20 μm. Fig. 2 View largeDownload slide Starch grain (SG) morphologies of three cereal species. (A–C) Seed images of rice (Oryza sativa), maize (Zea mays) and barley (Hordeum vulgare). Bars = 1 mm. (D–F) Iodine-stained sections of endosperm cells prepared using the rapid method. Bars = 20 μm. Boxes indicated by white dotted lines are illustrated areas in G–I. (G–I) Typical morphologies of SGs are illustrated. The SGs of rice are compound grains composed of many smaller granules. Maize SGs are the uniform type that are similar in size and round shaped. Barley SG is a bimodal type containing small and larger grains co-existing in the same cell. Red lines indicate cell walls. (J–L) SEM images. Bars = 20 μm. Screening of the mutants with distinct starch grain morphologies To isolate morphologically distinct SG mutants, we screened ethylmethane sulfonate (EMS)-treated M2 seeds from the lax panicle1 (lax1) mutant with a Nipponbare background using the rapid method. Because lax1 is a recessive mutation that results in the absence of lateral spikelets on panicles, sexual contamination could be monitored (Komatsu et al. 2001). We examined at least five seeds from each M2 line. Of the 1,152 M2 lines, five independent mutants with abnormal SG morphologies were isolated. All mutant grains showed a chalky endosperm (Fig. 3A). Subcellular images obtained by the rapid method showed that the morphologies of the parental lax1 and Nipponbare wild-type SGs were indistinguishable (Fig. 3B, Nipponbare lax1). This indicated that the lax1 mutation does not affect SG morphology. Normal SGs were uniform in size and approximately 10–20 μm in diameter. In ssg1, ssg2 and ssg3 mutants, many smaller, simple grains (<10 μm in diameter) were observed, in addition to the normal compound grains (Fig. 3B). In ssg4 mutants, larger SGs (>30 μm in diameter) were detected. SGs without internal compound structures were observed in ssg5 mutants. All of these phenotypes were inherited in subsequent generations. Fig. 3 View largeDownload slide Phenotypes of the isolated mutants. (A) Whole seeds (upper panel) and their cross-sections (lower panel) of the parental line (Nipponbare lax1) and mutants. Bars = 1 mm. (B) Iodine-stained endosperm sections prepared using the rapid method. Bars = 20 μm. Fig. 3 View largeDownload slide Phenotypes of the isolated mutants. (A) Whole seeds (upper panel) and their cross-sections (lower panel) of the parental line (Nipponbare lax1) and mutants. Bars = 1 mm. (B) Iodine-stained endosperm sections prepared using the rapid method. Bars = 20 μm. Starch grain morphologies of the ssg mutants observed using a conventional resin embedding method The rapid method described above allowed us to visualize SGs rapidly and easily. However, fine, thin sections were difficult to obtain from mutants with a floury endosperm, such as ssg4 and ssg5. To obtain fine images of SGs, we chemically fixed endosperm samples and embedded them in Technovit resin. Technovit sections (approximately 1 μm thick) were prepared and stained with iodine (Fig. 4). For the wild-type Nipponbare, the SG morphologies in the central and peripheral parts of the endosperm were similar to those obtained by the rapid method (Fig. 4A–D). In the central part of the ssg1 endosperm, most cells contained compound and small simple grains simultaneously (Fig. 4E, F), whereas the peripheral endosperm cells below the aleurone layer were filled with only compound grains (Fig. 4G, H). Fig. 4 View largeDownload slide Iodine-stained Technovit sections from ssg mutant endosperm. (A–D) Nipponbare, (E–H) ssg1, (I, J) ssg2, (K, L) ssg3, (M–P) ssg4, (Q–X) ssg5. The images shown in C, D, G, H, O, P, W and X are peripheral endosperm cells below the aleurone layer. Other images are from the central part of the endosperm. Asterisks indicate aleurone cells. In F, J and L, arrowheads and arrows indicate the positions of some small simple grains and compound grains, respectively. Bars = 20 μm. Fig. 4 View largeDownload slide Iodine-stained Technovit sections from ssg mutant endosperm. (A–D) Nipponbare, (E–H) ssg1, (I, J) ssg2, (K, L) ssg3, (M–P) ssg4, (Q–X) ssg5. The images shown in C, D, G, H, O, P, W and X are peripheral endosperm cells below the aleurone layer. Other images are from the central part of the endosperm. Asterisks indicate aleurone cells. In F, J and L, arrowheads and arrows indicate the positions of some small simple grains and compound grains, respectively. Bars = 20 μm. In the ssg2 and ssg3 mutants, the SG morphologies were quite similar to that of ssg1 (Fig. 4I–L). As in ssg1 mutants, the peripheral endosperm cells below the aleurone layer in ssg2 and ssg3 had only compound grains (data not shown). ssg4 SGs were so large that only a few SGs filled a single cell in the central part of the endosperm (Fig. 4M, N). This was also observed in the peripheral endosperm cells (Fig. 4O, P). In ssg5, SG morphologies varied between cells (Fig. 4Q–X). Some SGs stained weakly with iodine and their internal compound structures were not detected (Fig. 4R, S). Other SGs stained well with iodine, but their interior compound structures were abolished, especially in the central parts of the SGs (Fig. 4T–V). Abnormal SGs were most abundant in the central parts of endosperm cells, while normal-looking SGs were enriched in the peripheral endosperm cells below the aleurone layer (Fig. 4W, X). The structure of endosperm amylopectin in the mutant lines The chain length distribution of amylopectin is one of the molecular properties that determine amylopectin structure. Previously identified starch-related mutants often showed characteristic chain length distributions (Nakamura 2002). To evaluate the effect of ssg mutations on amylopectin structure, the chain length distributions [degree of polymerization (DP) ≤80] for isoamylolysate of endosperm amylopectin were analyzed using capillary electrophoresis (Fig. 5A–E). In ssg1 mutants, the proportion of short chains with DP ≤13 was markedly reduced, whereas the proportion of long chains with DP ≥38 was elevated (Fig. 5A). In particular, the intermediate length chains (14 ≤ DP ≤37) were significantly higher compared with the wild type. The same trend was observed in the ssg2 and ssg3 mutants (Fig. 5B, C). Compared with the wild type, no significant changes were detected in the ssg4 and ssg5 mutants (Fig. 5D, E). Fig. 5 View largeDownload slide Amylopectin structures and the accumulation of starch-branching enzymes in ssg mutants. (A–E) Chain length distributions of endosperm amylopectin from the parental line (black bars) and the ssg1, ssg2, ssg3, ssg4 and ssg5 mutants (represented by red, blue, green, purple and brown bars, respectively). The inset graphs compare the chain length distribution patterns (Δ molar %) between each mutant and the wild type. Values for the molar % and Δ molar % are the averages for three grains arbitrarily chosen from a single homozygous plant. Relative SDs of the molar % for each chain length from DP 4 to 80 were <0.3%. DP, degree of polymerization. (F) Immunoblot analysis using anti-BEIIb and anti-BEI antibodies. The proteins were extracted from the three mature seeds used for the chain length distribution analysis. Fig. 5 View largeDownload slide Amylopectin structures and the accumulation of starch-branching enzymes in ssg mutants. (A–E) Chain length distributions of endosperm amylopectin from the parental line (black bars) and the ssg1, ssg2, ssg3, ssg4 and ssg5 mutants (represented by red, blue, green, purple and brown bars, respectively). The inset graphs compare the chain length distribution patterns (Δ molar %) between each mutant and the wild type. Values for the molar % and Δ molar % are the averages for three grains arbitrarily chosen from a single homozygous plant. Relative SDs of the molar % for each chain length from DP 4 to 80 were <0.3%. DP, degree of polymerization. (F) Immunoblot analysis using anti-BEIIb and anti-BEI antibodies. The proteins were extracted from the three mature seeds used for the chain length distribution analysis. The chain length distribution patterns of ssg1, ssg2 and ssg3 amylopectin were quite similar to that of the previously identified ae mutant (Nishi et al. 2001, Nakamura 2002). In ae mutants, one of the starch-branching enzymes, BEIIb, is down-regulated (Mizuno et al. 1993). Starch-branching enzymes generate α-1,6 linkages by cleaving internal α-1,4 bonds and transferring the released, reducing ends to a C6 hydroxyl. Lack of BEIIb activity in ae results in abnormal amylopectin structures, which are enriched in long chains and depleted of short chains (Nishi et al. 2001). To investigate the accumulation of BEIIb in ssg mutants, total proteins were extracted from the mutant seeds and subjected to immunoblot analysis using anti-BEIIb antibodies (Fig. 5F). ssg1 seeds totally lacked BEIIb, whereas the other mutant seeds accumulated BEIIb to levels similar to the wild type. In contrast, the other starch-branching enzyme, BEI (Nakamura and Yamanouchi 1992), was expressed normally in all mutants. This indicates that, similar to the ae mutant, the abnormal amylopectin structure in ssg1 is caused by the lack of BEIIb. Genetic lesions in the BEIIb gene in the ssg1, ssg2 and ssg3 mutants When ssg1 was crossed with Kasalath, 20 out of 109 F2 seeds showed the co-existence of simple and compound grains in their endosperm, indicating that ssg1 segregated as a single recessive allele (χ2 = 2.6, P = 0.11). ssg1 was mapped to the middle of chromosome 2 where the BEIIb gene (Os02g0528200) is located (Harrington et al. 1997). Two splicing variants of the BEIIb gene were predicted based on expressed sequence tag (EST) analysis (LOC_Os02g32660.1 and LOC_Os02g32660.2, http://rice.plantbiology.msu.edu/). The shorter spliced variant is truncated relative to the longer transcript (Fig. 6A). We determined the genomic sequence of the longer BEIIb gene from the 5′ untranslated region (UTR) through to the 3′ UTR in the ssg1, ssg2 and ssg3 mutants, and found several base changes (Fig. 6A). In ssg1, two base changes were identified in introns. One was located in the 12th intron splicing acceptor site. The guanine residue (+5789) essential for correct mRNA splicing was replaced by adenine. The ssg2 mutant also had two base changes. One was located in the 18th exon, which caused an amino acid substitution from proline to leucine. In ssg3, a base change was found in the 17th exon, resulting in a glycine to arginine substitution. Fig. 6 View largeDownload slide Mutation sites in the BEIIb gene from the ssg1, ssg2 and ssg3 mutants. (A) Structure of the BEIIb gene (Os02g0528200). Two distinct splicing variants were predicted based on the EST analysis (LOC_Os02g32660.1, upper; LOC_Os02g32660.2, lower). The coding and untranslated regions are depicted as blue and white boxes, respectively. Introns are indicated by black bars. Adenine from the translation start codon (ATG) is designated as +1, and the stop codons (TGA) are +11,025 and +8,428, respectively. The ssg1 mutant had two base pair changes (G to A) at +5,789 and +10,703. The former change was located at the 5′ splicing acceptor site of the 12th intron. The ssg2 mutant also had two base pair changes (T to C and C to T) at +7,863 and +9,109, respectively. The latter change was located in the 18th exon and resulted in an amino acid substitution of Pro644 by leucine (Leu). The ssg3 mutant had a single base pair change (G to A) in the 17th exon, leading to an amino acid substitution of Gly667 by arginine (Arg). (B) The protein structure of the Oryza sativa BEIIb protein. Dark and light gray boxes represent the transit peptide and mature regions, respectively. Black boxes are the four catalytic sites conserved in amylolytic enzymes (region I, II, III and IV). The predicted positions of the β-strands and α-helices of the (β/α)8 barrel domain common to the amylolytic superfamily proteins are indicated with white and red boxes, respectively. Green and blue arrowheads indicate the mutation sites in ssg3 and ssg2, respectively. The open box is the region that was used for the alignment in C. (C) Alignment with other starch- and glycogen-branching enzymes. Enzymes indicated with blue and red characters belong to families A and B of the starch-branching enzymes, respectively. Glycogen-branching enzymes are indicated in black. OsBEIIb and OsBEI (Oryza sativa, 02g0528200 and Os06g0726400), ZmBEIIb and ZmBEI (Zea mays, NP_001105316 and NP_001105370), TaBEII and TaBEI (Triticum aestivum, Y11282 and Y12320), PsSBEII and PsSBEI (Pisum sativum, CAA56320 and CAA56319), StSBEII and StSBEI (Solanum tuberosum, CAB40748 and CAA49463), ScGLC3 (Saccharomyces cerevisiae, AAA34632) and HsGBEI (Homo sapiens, NM_000158). Perfectly conserved residues are shown in black. Blue and green arrowheads indicate the residues substituted by ssg3 and ssg2 mutations, respectively. The Clustal X program (Larkin et al. 2007) was used for the alignment. Fig. 6 View largeDownload slide Mutation sites in the BEIIb gene from the ssg1, ssg2 and ssg3 mutants. (A) Structure of the BEIIb gene (Os02g0528200). Two distinct splicing variants were predicted based on the EST analysis (LOC_Os02g32660.1, upper; LOC_Os02g32660.2, lower). The coding and untranslated regions are depicted as blue and white boxes, respectively. Introns are indicated by black bars. Adenine from the translation start codon (ATG) is designated as +1, and the stop codons (TGA) are +11,025 and +8,428, respectively. The ssg1 mutant had two base pair changes (G to A) at +5,789 and +10,703. The former change was located at the 5′ splicing acceptor site of the 12th intron. The ssg2 mutant also had two base pair changes (T to C and C to T) at +7,863 and +9,109, respectively. The latter change was located in the 18th exon and resulted in an amino acid substitution of Pro644 by leucine (Leu). The ssg3 mutant had a single base pair change (G to A) in the 17th exon, leading to an amino acid substitution of Gly667 by arginine (Arg). (B) The protein structure of the Oryza sativa BEIIb protein. Dark and light gray boxes represent the transit peptide and mature regions, respectively. Black boxes are the four catalytic sites conserved in amylolytic enzymes (region I, II, III and IV). The predicted positions of the β-strands and α-helices of the (β/α)8 barrel domain common to the amylolytic superfamily proteins are indicated with white and red boxes, respectively. Green and blue arrowheads indicate the mutation sites in ssg3 and ssg2, respectively. The open box is the region that was used for the alignment in C. (C) Alignment with other starch- and glycogen-branching enzymes. Enzymes indicated with blue and red characters belong to families A and B of the starch-branching enzymes, respectively. Glycogen-branching enzymes are indicated in black. OsBEIIb and OsBEI (Oryza sativa, 02g0528200 and Os06g0726400), ZmBEIIb and ZmBEI (Zea mays, NP_001105316 and NP_001105370), TaBEII and TaBEI (Triticum aestivum, Y11282 and Y12320), PsSBEII and PsSBEI (Pisum sativum, CAA56320 and CAA56319), StSBEII and StSBEI (Solanum tuberosum, CAB40748 and CAA49463), ScGLC3 (Saccharomyces cerevisiae, AAA34632) and HsGBEI (Homo sapiens, NM_000158). Perfectly conserved residues are shown in black. Blue and green arrowheads indicate the residues substituted by ssg3 and ssg2 mutations, respectively. The Clustal X program (Larkin et al. 2007) was used for the alignment. To assess the linkage between the base change and the ssg3 SG phenotype, we next designed the derived cleaved-amplified polymorphic sequence (dCAPS) primers according to the base change in ssg3 (+8,898 in Fig. 6A). The dCAPS successfully detected the genotypes of wild-type, ssg3 and F1 heterozygous plants (Supplementary Fig. S1). When ssg3 was crossed with Kasalath, 21 out of 79 F2 seeds showed the SG phenotype specific to ssg3, indicating that ssg3 is single recessive (χ2 = 0.11, P = 0.75). Genotyping of the 21 mutant seeds using the dCAPS primer showed that all 21 seeds were homozygous for the ssg3 base change (Supplementary Fig. S1). This result supports the idea that the base change in the BEIIb gene in the ssg3 mutant is responsible for the ssg3 SG phenotype. Based on the structural analysis of other glucanases, branching enzymes are predicted to contain a catalytic (β/α)8 barrel domain that acts as a scaffold for substrate binding and catalysis (Buisson et al. 1987, Jespersen et al. 1993, Abad et al. 2002). The substituted amino acids in ssg2 and ssg3 were located inside the loop between the eighth β-strand and the α-helix, and outside the conserved catalytic domain (regions I, II, III and IV, Fig. 6B) (Svensson 1994). Plant starch-branching enzymes are separated into two families, A and B, based on their amino acid sequence similarities (Burton et al. 1995, Mizuno et al. 2001). Rice BEIIb belongs to family A and BEI is a member of family B. The proline residue substituted by the ssg2 mutation is conserved in both families, while the glycine residue substituted in ssg3 is conserved only in family A (Fig. 6C). Discussion Usefulness of the rapid method for preparing thin sections of endosperm It has been well documented that SG morphologies show significant differences between plant species (Harz 1880, Tateoka 1954, Tateoka 1955, Tateoka 1962, Czaja 1978, Jane et al. 1994, James et al. 2003, Shapter et al. 2008). Various morphologies of SGs in many plant species have been observed in extracted samples; however, subcellular SGs have rarely been observed except for those found in major cereal species. In this study, we developed a method to prepare thin sections of cereal endosperm (Fig. 1) that allowed us to visualize the subcellular SGs in endosperm cells rapidly and easily (Fig. 2). Another advantage of this method was that seeds used for sectioning remained alive and ready to grow because chemical fixation was not required. This enabled easier genetic screening based on the direct observation of SGs (Fig. 3). Using this method, we found that >200 rice seeds could be examined in a single day (data not shown). In addition, for wild-type seeds, the images obtained using the rapid method were quite similar to those obtained by conventional resin sectioning methods (Figs. 3B, 4A–D). However, in seeds with floury properties, such as those from ssg4 and ssg5, fine, thin sections were hard to prepare using the rapid method. Therefore, the images of sections obtained by the rapid method were lower in quality than for sections obtained using resin sectioning methods (Figs. 3B, 4M–X). Possible allelic relationship between ssg1, ssg2, ssg3 and ae mutants Out of five ssg mutants isolated in this study, three (ssg1, ssg2 and ssg3) had mutations in the BEIIb gene (Fig. 6). Rice ae mutants were originally isolated because of their increased amylose content (Yano et al. 1985). Further analysis revealed that the amylopectin structure, not amylose levels, is affected in ae, due to the lack of the BEIIb protein (Mizuno et al. 1993, Nishi et al. 2001). The mutation sites of rice ae have not yet been determined. However, the genomic region covering the BEIIb gene can complement the altered structure of ae amylopectin, suggesting that defective BEIIb is responsible for ae (Tanaka et al. 2004). ae SGs observed by SEM display irregular, round-shaped morphologies (Yano et al. 1985, Tanaka et al. 2004). The number of round-shaped simple grains increased dramatically in ssg1, ssg2 and ssg3 (Fig. 4E, F, I–L). In addition, the amylopectin chain length distributions (Fig. 5) were very similar in ssg1, ssg2, ssg3 and ae mutants. Together, these data suggest a possible allelic relationship between ssg1, ssg2, ssg3 and ae. In ssg1, the point mutation (+5,789) was located at a guanine residue at the splicing acceptor site (Fig. 6A). Thus, a lack of mature mRNAs probably leads to a significant reduction in BEIIb protein levels (Fig. 5F). In ssg2 and ssg3, point mutations (+9,109 and +8,898, respectively) are expected to cause amino acid substitutions; however, these mutations had no effect on the accumulation of the BEIIb protein (Fig. 5F). Previous domain-swapping experiments between maize BEI and BEII showed that the C-terminal domain of BE is involved in substrate specificity (Kuriki et al. 1997). ssg2 and ssg3 mutations were located in the C-terminal domain used in the domain-swapping experiments. Collectively, these data suggest that the amino acids substituted in ssg2 and ssg3 are crucial for BEIIb activity, and that the accumulated BEIIb proteins in ssg2 and ssg3 are enzymatically inactive. ssg4 and ssg5 will provide new insights into amyloplast division and starch grain morphologies Genetic mapping of the ssg4 and ssg5 loci is now underway. Therefore, we have not yet determined whether they are novel mutants. Out of the previously isolated starch-related mutants, sugary-1 and floury 2 (flo2) mutants have some common characteristics with ssg4 and ssg5. ssg4 and ssg5 seeds showed floury phenotypes (data not shown) and some cells in ssg5 were stained weakly with iodine (Fig. 4R), which is characteristic of sugary-1 mutants (Nakamura et al. 1996). sugary-1 lacks one of the starch-debranching enzymes, isoamylase 1, which leads to the production of highly and randomly branched polyglucan (phytoglycogen) instead of amylopectin (Nakamura et al. 1997). Technovit sections from sugary-1 endosperm showed that some SGs in the central part of the endosperm were stained weakly by iodine and were strikingly amorphous (Supplementary Fig. S2A–C). Consistent with previous observations (Kawagoe et al. 2005), compound grains consisting of much smaller starch granules were also detected (Supplementalry Fig. S2D). In addition to the abnormal SGs, normal SGs were also observed in the central and peripheral parts of the sugary-1 endosperm (Supplementary Fig. S2E–H). In the case of flo2, SGs showed normal morphologies and sizes (Supplementary Fig. S2I, J), indicating that ssg4 and ssg5 are different from flo2. The morphological variations of SGs and the presence of cells that stain weakly with iodine are commonly observed in ssg5 and sugary-1 (Fig. 4Q–S and Supplementary Fig. S2A–C); however, they appear to be different mutants for the following reasons. In sugary-1, the phytoglycogen-concentrated region that is not stained with iodine dominates a large portion of the central part of the endosperm (Nakamura et al. 1997), whereas the weakly stained region was not obvious in ssg5 under a stereo microscope (Supplementary Fig. S3). Furthermore, the SGs that contained a simple globular structure in the central part, as shown in Fig. 4 (Q and U), were specific to ssg5. The chain length distribution of amylopectin in ssg5 and the wild type was quite similar (Fig. 5E), suggesting that the gene responsible for ssg5 is not related to amylopectin biosynthesis. In contrast, sugary-1 amylopectin showed significant increases in the number of short chains (DP ≤10), and a depletion of intermediate (10≤ DP ≤30) and long chains (40≤ DP), because of a lack of the starch-debranching enzyme, isoamylase 1 (Kubo et al. 1999). Therefore, we concluded that ssg5 is a different mutation from sugary-1 and may be a novel factor required to determine or maintain SG morphologies. The ssg4 SGs were an enlarged size, while the cell size was significantly smaller compared with the wild type (Fig. 4A–D, M–P). Recent studies using Arabidopsis mutants have demonstrated that inhibition of chloroplast division causes chloroplast enlargement (Glynn et al. 2007). For example, the Arabidopsis mutant arc5 (accumulation and replication of chloroplast 5) has only three chloroplasts in a single cell (Gao et al. 2003). ARC5 encodes a member of the dynamin superfamily, which plays a central role in chloroplast division (Gao et al. 2003). Similarly, the division of amyloplasts may be inhibited in ssg4 endosperm. Rice arc5 mutants have been isolated and reported to have enlarged chloroplasts (Yun and Kawagoe 2009). However, rice arc5 mutants develop pleomorphic amyloplasts that are sometimes elongated and have a beads-on-a-string structure (Yun and Kawagoe 2009). The morphologies of ssg4 and arc5 amyloplasts are different from each other, suggesting that ssg4 is distinct from arc5. Different division mechanisms between chloroplasts and amyloplasts have been proposed (Yun and Kawagoe 2009). In leaves, chloroplasts divide by binary fission, in which fission of the thylakoid membrane precedes the separation of the envelope membrane. During binary fission, dividing chloroplasts often adopt dumbbell-like shapes (Glynn et al. 2007). In amyloplasts, two different division mechanisms have been suggested: (i) the multiple, beads-on-a-string-type; and (ii) the budding-type. In multiple, beads-on-a-string-type division, elongated amyloplasts generate multiple constriction sites to which ARC5 and another chloroplast division-related factor, FtsZ, are recruited. The beads-on-a-string structure is observed when the endosperm cell is still small and rapidly expanding in volume. At later stages of endosperm development, large amyloplasts divide by the protrusion of smaller amyloplasts from their surfaces. This protrusion process is called budding-type division and is achieved by unknown mechanisms. SSG4 may encode a novel factor that has not been discovered by previous, chloroplast-focused studies, and that may be involved in amyloplast budding-type division. Morphological variations of starch grains between peripheral and central parts of the endosperm SG morphologies affected by ssg mutations varied between peripheral and central parts of the endosperm (Fig. 4). In ssg mutants, except for ssg4, most SGs in the peripheral endosperm cells looked normal, while abnormal SGs were enriched in the central portion of the endosperm. This pattern was also observed in the sugary-1 mutant (Supplementary Fig. S2A–H), suggesting the general character of SG morphologies. In rice, SG formation begins in the innermost cells of the endosperm and spreads to outer cells centrifugally. The peripheral endosperm cells are the last parts to be filled with SGs and this process is completed about 1 month after the start of SG formation in the central region (Hoshikawa 1993). Gene expression of starch-related enzymes changes dramatically between the early and late stages of seed development (Ohdan et al. 2005). Therefore, the molecular mechanisms responsible for SG formation may also change within this time period. Detailed microscopic observation of SGs during all stages of endosperm development in ssg mutants is necessary for a better understanding of SG formation. The endosperm of all ssg mutants showed chalkiness that was limited to the central portions (Fig. 3A). Generally, the chalky parts were not sufficiently filled with SGs and had many minute gaps that caused light to scatter within these regions, giving them a white appearance (Hoshikawa 1993). Abnormal SG morphologies in ssg mutants may lead to cells that were incompletely filled with SGs and cause the chalky appearance, especially in the central part of the endosperm. However, in rare cases of ssg1, non-chalky, translucent grains were also obtained (Supplementary Fig. S4A, B) where SG morphologies were indistinguishable from those in the chalky ssg1 endosperm (Supplementary Fig. S4C). This suggests that the abnormal SG morphologies are not the direct cause of chalkiness in ssg1. Concluding remarks Many Poaceae species have been examined, mostly in extracted conditions, for SG morphologies (Tateoka 1962). We are now re-evaluating SG morphologies of the Poaceae species using the rapid and resin-embedding methods to verify previous observations and discover novel types of SGs with unique morphologies. Molecular analysis focusing on the mechanisms that determine SG morphology and descriptive analysis of cross-species SG diversity will lead to a more complete understanding of the molecular diversity of SG morphologies. The methods reported in this study should serve as an effective technique for these analyses. Materials and Methods Plant material and growth The endosperm mutant lines used in this experiment were selected from lax1 mutants that were originally obtained from regenerated plants derived from Nipponbare calli. Mutagenization was carried out by soaking lax1 seeds in 1.5% (v/v) methanesulfonic acid ethyl ester (Sigma, Tokyo, Japan). The M2 line derived from a single M1 plant were grown and M2 seeds were collected from individual M1 plants after self-fertilization. Rice plants were grown at an experimental field of the Institute of Plant Science and Resources, Okayama University under natural conditions or at 28°C in a greenhouse. Seeds of Zea mays (maize) were purchased from Nagano-ken Yuukiseisansharengou Ltd. (Nagano, Japan). Seeds of the sugary-1 mutant (Tankei 2013) were provided by the Genebank of the National Institute of Agrobiological Sciences. floury 2 mutants were previously isolated from the γ-ray-treated M2 population (Maekawa 1985). Isolation of starch grain mutants by the rapid observation method Screening was carried out with at least five seeds from each M2 line (1,152 lines). Endosperm thin sections from the M2 seeds were prepared by the rapid method. A mature M2 seed was inserted into a truncated pipet tip (Fig. 1A, 200 μl pipet tip, #110, Quality Scientific Plastics, Petaluma, CA, USA). The seed-captured tip was fixed on the block trimmer (Fig. 1B, C; Okenshoji, Tokyo, Japan) that was originally developed for resin block trimming for ultramicrotomy. The fixed seeds on the block trimmer were manipulated under a stereo microscope (Fig. 1D, SZ61, Olympus, Tokyo, Japan) and were trimmed with a razor blade (#FH-10, Feather anzen kamisori, Osaka, Japan) to generate a smooth surface on the top of the seed. The positions of the hands, fingers, a razor blade and the block trimmer during trimming and thin sectioning are described in the Results section. Starch staining was done by immersion in a drop of deionized water containing 40 times diluted Lugol solution (iodine/potassium iodine solution, MP Biomedicals, Eschwege, Germany). The samples were subsequently examined under a microscope (AX70, Olympus). Scanning electron microscopy To observe endosperm starch grains by SEM, dry seeds of rice, maize and barley were cut with a razor blade to expose the fractured surfaces of the endosperm. The specimens were coated with gold by a sputter coating machine (MSP-1S, Shinkuu device, Ibaraki, Japan) and examined with SEM (Quanta 250, FEI, Hillsboro, OR, USA). Thin sections of Technovit 7100 resin-embedded endosperm Approximately 1 mm cubic blocks were cut out from the endosperm of dry seeds and fixed in FAA solution containing 5% (v/v) formalin, 5% (v/v) acetic acid and 50% (v/v) ethanol, for at least 12 h at room temperature. Samples were subsequently dehydrated through a graded ethanol series [30, 50, 70, 90 and 100% (v/v)] and then embedded in Technovit 7100 resin (Kulzer and Company, Wehrheim, Germany). The embedded samples were cut with an Ultracut N ultramicrotome (Reichert-Nissei, Tokyo, Japan) and glass knives, and dried on coverslips. Thin sections (approximately 1 μm thickness) were stained with 40 times diluted Lugol solution in deionized water for at least 5 s and subsequently examined under a microscope (AX70). Chain length distribution of endosperm amylopectin by capillary electrophoresis To extract starch from mature endosperm for amylopectin chain length distribution, embryo removed from mature dry seeds was crushed with pliers and hand homogenized using a mortar and pestle. It was then suspended in 5 ml of methanol and boiled for 10 min. The homogenate was centrifuged at 2,500 × g for 10 min. The precipitate was washed twice with 5 ml of 90% (v/v) methanol and suspended in 300 μl of 0.25 M NaOH. The suspension was boiled for 5 min. A 9.6 μl aliquot of 100% acetic acid, 100 μl of 2% (w/v) NaN3 and 1,090 μl of distilled water were added to the gelatinized α-polyglucan sample. The sample was hydrolyzed by adding 4 μl of Pseudomonas amyloderamosa isoamylase (354 U; Hayashibara, Okayama, Japan) at 37°C for 24 h. The hydrolyzed sample was boiled for 20 min and centrifuged. The supernatant was deionized by filtration on an ion exchange resin [BioRad AG501-X8(D)] in a microtube. An appropriate aliquot containing 10 nmol of reduced end estimated by the modified Park Johnson method (Hizukuri et al. 1981) was evaporated to dryness in a centrifugal vacuum evaporator (Taitec, Tokyo, JApan). Fluorescence labeling and capillary electrophoresis were performed according to the previously reported method (O’Shea and Morell 1996) and the protocols provided by the manufacturer by using the eCAP N-linked oligosaccharide profiling kit and capillary electrophoresis (P/ACE MDQ Capillary Electrophoresis System; Beckman Coulter’s, Fullerton, CA, USA). Immunoblot analysis The rice flours were prepared from each mutant and the parental line and were mixed in 50 vols. of extraction buffer: 8 M urea, 5% (v/v) 2-mercaptoethanol, 4% (w/v) SDS, 125 mM Tris–HCl, pH 6.8. The samples were incubated at 37°C for 12 h and subjected to centrifugation at 20,000 × g at 20°C for 10 min. The 10 μl supernatant was mixed with 5 μl of SDS solution containing 55 mM Tris–HCl (pH 6.8), 2.3% SDS, 5% 2-mercaptoethanol and 10% glycerol, and subjected to 7.5% SDS–PAGE. After SDS–PAGE, the proteins were transferred electrophoretically to a polyvinylidene difuoride (PVDF) membrane (Millipore, Tokyo, Japan) by transblotter (Nihon Eido Co., Tokyo, Japan). The membrane was then incubated in Tris-buffered saline (pH 7.5) plus 0.05% (v/v) Tween-20 with the anti-BEIIb and anti-BEI antibodies (Nakamura et al. 1992) for 1 h. Dilutions of the antibodies are 1 : 1,000 (v/v). Horseradish peroxidase-conjugated goat antibodies against rabbit IgG (Bio-Rad, Hercules, CA, USA) were diluted (1 : 2,000) to be used as second antibodies. The immunoreactive bands were finally detected with 4-chloro-1-naphthol. Mapping of the SSG1 locus and determination of the mutation site of ssg1, ssg2 and ssg3 mutants For mapping of the SSG1 gene, we constructed an F2 population derived from a cross between the ssg1 mutant and Kasalath. Endosperm thin sections of each F2 seed were examined by the rapid method to select ssg1 mutant seeds. Rough mapping of this gene was done using 24 F2ssg1 plants. The genomic DNA of these ssg1 mutants was individually isolated and analyzed using simple sequence length polymorphism markers (Temnykh et al. 2000, McCouch et al. 2002, Maekawa et al. 2005) to determine the molecular markers linked to SSG1. SSG1 was defined to the region where BEIIb gene was located. BEIIb gene regions covering from the 5′ UTR to the 3′ UTR of ssg1, ssg2 and ssg3 were sequenced using the BigDye Terminators v1.1 cycle sequencing kit (Applied Biosystems, Foster City, CA, USA) and a 3100-Avant Genetic Analyzer (Applied Biosystems). Co-segregation analysis of the base change in the BEIIb gene with the ssg3 phenotype To confirm that the base change (+8,898 in Fig. 6A) in the ssg3 mutant co-segregates with the ssg phenotype, we crossed ssg3 with Kasalath and produced F2 seed populations. Out of the 79 F2 seeds, 21 ssg seeds were selected by the rapid method and genomes were isolated. To detect the base change in ssg3, dCAPS primers were designed as follows; 5′-TGAATTT TAATGAATACTGGCATGA-3′and 5′-AAAGTTAAGATAGCC TTCTCCTCCTGAGC-3′. The PCR conditions were as follows: 94°C for 2 min, and 35 cycles of 94°C for 30 s, 50°C for 45 s and 68°C for 1 min. The PCR product was digested with SacI and PCR products were subsequently separated by 15% PAGE and detected with ethidium bromide staining. In the case of ssg3, a PCR product (174 bp) was digested into 148 and 26 bp. In the case of the wild type (Nipponbare), the PCR product (174 bp) was not digested. Funding The Ministry of Education, Culture, Sports, Science and Technology [Grant-in-Aid for Scientific Research (No. 20770036 to R.M.)]; the Iijima Memorial Foundation for the Promotion of Food Science and Technology; the Science and Technology Foundation of Japan; the Oohara Foundation. Acknowledgments The authors would like to thank Dr. Shin Taketa (Institute of Plant Science and Resources, Okayama University) for providing the barley seeds, and Dr. Jun Yamashita (Institute of Plant Science and Resources, Okayama University) for critical reading of the manuscript and valuable suggestions. The seeds of sugary-1 mutant (Tankei 2013) were kindly provided by the Genebank of the National Institute of Agrobiological Sciences (Tsukuba, Japan). We would also like to thank Rie Hijiya and Rumiko Itoh for their technical assistance. Abbreviations Abbreviations ae amylose-extender BE branching enzyme dCAPS derived cleaved-amplified polymorphic sequence DP degree of polymerization EST expressed sequence tag SEM scanning electron microscopy SG starch grain ssg substandard starch grain UTR untranslated region. References Abad MC,  Binderup K,  Rios-Steiner J,  Arni RK,  Preiss J,  Geiger JH.  The X-ray crystallographic structure of Escherichia coli branching enzyme,  J. Biol. Chem. ,  2002, vol.  277 (pg.  42164- 42170) Google Scholar CrossRef Search ADS PubMed  Buisson G,  Duee E,  Haser R,  Payan F.  Three dimensional structure of porcine pancreatic α-amylase at 2.9 Å resolution. Role of calcium in structure and activity,  EMBO J. ,  1987, vol.  6 (pg.  3909- 3916) Google Scholar PubMed  Buléon A,  Colonna P,  Planchot V,  Ball S.  Starch granules: structure and biosynthesis,  Int. J. Biol. Macromol. ,  1998, vol.  23 (pg.  85- 112) Google Scholar CrossRef Search ADS PubMed  Burton RA,  Bewley JD,  Smith AM,  Bhattacharyya MK,  Tatge H,  Ring S, et al.  Starch branching enzymes belonging to distinct enzyme families are differentially expressed during pea embryo development,  Plant J. ,  1995, vol.  7 (pg.  3- 15) Google Scholar CrossRef Search ADS PubMed  Burton RA,  Jenner H,  Carrangis L,  Fahy B,  Fincher GB,  Hylton C, et al.  Starch granule initiation and growth are altered in barley mutants that lack isoamylase activity,  Plant J. ,  2002, vol.  31 (pg.  97- 112) Google Scholar CrossRef Search ADS PubMed  Czaja AT.  Structure of starch grains and the classification of vascular plant families,  Taxon ,  1978, vol.  27 (pg.  463- 470) Google Scholar CrossRef Search ADS   Dvonch W,  Kramer HH,  Whistler RL.  Polysaccharides of high-amylose corn,  Cereal Chem. ,  1951, vol.  28 (pg.  270- 280) Fujita N,  Kubo A,  Suh D-S,  Wong K-S,  Jane J-L,  Ozawa K, et al.  Antisense inhibition of isoamylase alters the structure of amylopectin and the physicochemical properties of starch in rice cndosperm,  Plant Cell Physiol. ,  2003, vol.  44 (pg.  607- 618) Google Scholar CrossRef Search ADS PubMed  Fujita N,  Toyosawa Y,  Utsumi Y,  Higuchi T,  Hanashiro I,  Ikegama A, et al.  Characterization of pullulanase (PUL)-deficient mutants of rice (Oryza sativa L.) and the function of PUL on starch biosynthesis in the developing rice endosperm,  J. Exp. Bot. ,  2009, vol.  60 (pg.  1009- 1023) Google Scholar CrossRef Search ADS PubMed  Fujita N,  Yoshida M,  Kondo T,  Saito K,  Utsumi Y,  Tokunaga T, et al.  Characterization of SSIIIa-deficient mutants of rice: the function of SSIIIa and pleiotropic effects by SSIIIa deficiency in the rice endosperm,  Plant Physiol. ,  2007, vol.  144 (pg.  2009- 2023) Google Scholar CrossRef Search ADS PubMed  Gao H,  Kadirjan-Kalbach D,  Froehlich JE,  Osteryoung KW.  ARC5, a cytosolic dynamin-like protein from plants, is part of the chloroplast division machinery,  Proc. Natl Acad. Sci. USA ,  2003, vol.  100 (pg.  4328- 4333) Google Scholar CrossRef Search ADS   Glynn JM,  Miyagishima S-Y,  Yoder DW,  Osteryoung KW,  Vitha S.  Chloroplast division,  Traffic ,  2007, vol.  8 (pg.  451- 461) Google Scholar CrossRef Search ADS PubMed  Hancock RD,  Tarbet BJ.  The other double helix—the fascinating chemistry of starch,  J. Chem. Educ. ,  2000, vol.  77 (pg.  988- 992) Google Scholar CrossRef Search ADS   Harrington SE,  Bligh HFJ,  Park WD,  Jones CA,  McCouch SR.  Linkage mapping of starch branching enzyme III in rice (Oryza sativa L.) and prediction of location of orthologous genes in other grasses,  Theor. Appl. Genet. ,  1997, vol.  94 (pg.  564- 568) Google Scholar CrossRef Search ADS   Harz CO.  Beiträge zur systematik der Gramineen,  Linnaea ,  1880, vol.  43 (pg.  1- 30) Hizukuri S,  Takeda Y,  Yasuda M,  Suzuki A.  Multi-branched nature of amylose and the action of debranching enzymes,  Carbohydr. Res. ,  1981, vol.  94 (pg.  205- 213) Google Scholar CrossRef Search ADS   Hoshikawa K.  Matsuo T,  Hoshikawa K.  Anthesis, fertilization and development of caryopsis,  Science of the Rice Plant. Vol. 1 Morphology ,  1993 Tokyo Food and Agriculture Policy Research Center(pg.  339- 376) James MG,  Denyer K,  Myers AM.  Starch synthesis in the cereal endosperm,  Curr. Opin. Plant Biol. ,  2003, vol.  6 (pg.  215- 222) Google Scholar CrossRef Search ADS PubMed  Jane J-L,  Kasemsuwan T,  Leas S,  Ames IA,  Zobel H,  Darien IL, et al.  Anthology of starch granule morphology by scanning electron microscopy,  Starch/Stärke ,  1994, vol.  46 (pg.  121- 129) Google Scholar CrossRef Search ADS   Jarvi AJ,  Eslick RF.  Shrunken endosperm mutants in barley,  Crop Sci. ,  1975, vol.  15 (pg.  363- 366) Google Scholar CrossRef Search ADS   Jespersen HM,  MacGregor EA,  Henrissat B,  Sierks MR,  Svensson B.  Starch- and glycogen-debranching and branching enzymes: prediction of structural features of the catalytic (β/α)8-barrel domain and evolutionary relationship to other amylolytic enzymes,  J. Protein Chem. ,  1993, vol.  12 (pg.  791- 805) Google Scholar CrossRef Search ADS PubMed  Kang HG,  Park S,  Matsuoka M,  An G.  White-core endosperm floury endosperm-4 in rice is generated by knockout mutations in the C4-type pyruvate orthophosphate dikinase gene (OsPPDKB),  Plant J. ,  2005, vol.  42 (pg.  901- 911) Google Scholar CrossRef Search ADS PubMed  Kawagoe Y,  Kubo A,  Satoh H,  Takaiwa F,  Nakamura Y.  Roles of isoamylase and ADP-glucose pyrophosphorylase in starch granule synthesis in rice endosperm,  Plant J. ,  2005, vol.  42 (pg.  164- 174) Google Scholar CrossRef Search ADS PubMed  Komatsu M,  Maekawa M,  Shimamoto K,  Kyozuka J.  The LAX1 and FRIZZY PANICLE 2 genes determine the inflorescence architecture of rice by controlling rachis-branch and spikelet development,  Dev. Biol. ,  2001, vol.  231 (pg.  364- 373) Google Scholar CrossRef Search ADS PubMed  Kubo A,  Fujita N,  Harada K,  Matsuda T,  Satoh H,  Nakamura Y.  The starch-debranching enzymes isoamylase and pullulanase are both involved in amylopectin biosynthesis in rice endosperm,  Plant Physiol. ,  1999, vol.  121 (pg.  399- 409) Google Scholar CrossRef Search ADS PubMed  Kuriki T,  Stewart DC,  Preiss J.  Construction of chimeric enzymes out of maize endosperm branching enzymes I and II: activity and properties,  J. Biol. Chem. ,  1997, vol.  272 (pg.  28999- 29004) Google Scholar CrossRef Search ADS PubMed  Larkin MA,  Blackshields G,  Brown NP,  Chenna R,  McGettigan PA,  McWilliam H, et al.  Clustal W and Clustal X version 2.0,  Bioinformatics ,  2007, vol.  23 (pg.  2947- 2948) Google Scholar CrossRef Search ADS PubMed  Li J-H,  Guiltinan MJ,  Thompson DB.  Mutation of the maize sbe1a and ae genes alters morphology and physical behavior of wx-type endosperm starch granules,  Carbohydr. Res. ,  2007, vol.  342 (pg.  2619- 2627) Google Scholar CrossRef Search ADS PubMed  Maekawa M.  Location of a floury endosperm gene in the second linkage group,  Rice Genet. Newslett. ,  1985, vol.  2 (pg.  57- 58) Maekawa M,  Takamure I,  Ahmed N,  Kyozuka J.  Bunketsu-waito, one of the tillering dwarf, is controlled by a single recessive gene in rice (Oryza sativa L.),  Breed. Sci. ,  2005, vol.  55 (pg.  193- 196) Google Scholar CrossRef Search ADS   McCouch SR,  Teytelman L,  Xu Y,  Lobos KB,  Clare K,  Walton M, et al.  Development and mapping of 2240 new SSR markers for rice (Oryza sativa L.),  DNA Res. ,  2002, vol.  9 (pg.  199- 207) Google Scholar CrossRef Search ADS PubMed  Mizuno K,  Kawasaki T,  Shimada H,  Satoh H,  Kobayashi E,  Okumura S, et al.  Alteration of the structural properties of starch components by the lack of an isoform of starch branching enzyme in rice seeds,  J. Biol. Chem. ,  1993, vol.  268 (pg.  19084- 19091) Google Scholar PubMed  Mizuno K,  Kobayashi E,  Tachibana M,  Kawasaki T,  Fujimura T,  Funane K, et al.  Characterization of an isoform of rice starch branching enzyme, RBE4, in developing seeds,  Plant Cell Physiol. ,  2001, vol.  42 (pg.  349- 357) Google Scholar CrossRef Search ADS PubMed  Nakamura Y.  Towards a better understanding of the metabolic system for amylopectin biosynthesis in plants: rice endosperm as a model tissue,  Plant Cell Physiol. ,  2002, vol.  43 (pg.  718- 725) Google Scholar CrossRef Search ADS PubMed  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 mutant of rice,  Plant J. ,  1997, vol.  12 (pg.  143- 153) Google Scholar CrossRef Search ADS   Nakamura Y,  Takeichi T,  Kawagushi K,  Yamanouchi H.  Purification of two forms of starch branching enzyme (Q-enzyme) from developing rice endosperm,  Physiol. Plant. ,  1992, vol.  84 (pg.  329- 335) Google Scholar CrossRef Search ADS   Nakamura Y,  Umemoto T,  Takahata Y,  Komae K,  Amano E,  Satoh H.  Changes in structure of starch and enzyme activities affected by sugary mutations in developing rice endosperm. Possible role of starch debranching enzyme (R-enzyme) in amylopectin biosynthesis,  Physiol. Plant. ,  1996, vol.  97 (pg.  491- 498) Google Scholar CrossRef Search ADS   Nakamura Y,  Yamanouchi H.  Nucleotide sequence of a cDNA encoding starch-branching enzyme, or Q-enzyme I, from rice endosperm,  Plant Physiol. ,  1992, vol.  99 (pg.  1265- 1266) Google Scholar CrossRef Search ADS PubMed  Nishi A,  Nakamura Y,  Tanaka N,  Satoh H.  Biochemical and genetic analysis of the effects of Amylose-Extender mutation in rice endosperm,  Plant Physiol. ,  2001, vol.  127 (pg.  459- 472) Google Scholar CrossRef Search ADS PubMed  O’Shea MG,  Morell MK.  High resolution slab gel electrophoresis of 8-amino-1,3, 6-pyrenetrisulfonic acid (APTS) tagged oligosaccharides using a DNA sequencer,  Electrophoresis ,  1996, vol.  17 (pg.  681- 686) Google Scholar CrossRef Search ADS PubMed  Ohdan T,  Francisco PBJr,  Sawada T,  Hirose T,  Terao T,  Satoh H, et al.  Expression profiling of genes involved in starch synthesis in sink and source organs of rice,  J. Exp. Bot. ,  2005, vol.  56 (pg.  3229- 3244) Google Scholar CrossRef Search ADS PubMed  Satoh H,  Nishi A,  Fujita N,  Kubo A,  Nakamura Y,  Kawasaki T, et al.  Isolation and characterization of starch mutants in rice,  J. Appl. Glycosci. ,  2003, vol.  50 (pg.  225- 230) Google Scholar CrossRef Search ADS   Satoh H,  Nishi A,  Yamashita K,  Takemoto Y,  Tanaka Y,  Hosaka Y, et al.  Starch-branching enzyme I-deficient mutation specifically affects the structure and properties of starch in rice endosperm,  Plant Physiol. ,  2003, vol.  133 (pg.  1111- 1121) Google Scholar CrossRef Search ADS PubMed  Satoh H,  Omura T.  New endosperm mutations induced by chemical mutagenesis in rice, Oryza sativa L,  Jpn. J. Breed. ,  1981, vol.  31 (pg.  316- 326) Google Scholar CrossRef Search ADS   Satoh H,  Shibahara K,  Tokunaga T,  Nishi A,  Tasaki M,  Hwang SK, et al.  Mutation of the plastidial α-glucan phosphorylase gene in rice affects the synthesis and structure of starch in the endosperm,  Plant Cell ,  2008, vol.  20 (pg.  1833- 1849) Google Scholar CrossRef Search ADS PubMed  Shapter FM,  Henry RJ,  Lee LS.  Endosperm and starch granule morphology in wild cereal relatives,  Plant Genet. Resour. ,  2008, vol.  6 (pg.  85- 97) Google Scholar CrossRef Search ADS   Svensson B.  Protein engineering in the α-amylase family: catalytic mechanism, substrate specificity, and stability,  Plant Mol. Biol. ,  1994, vol.  25 (pg.  141- 157) Google Scholar CrossRef Search ADS PubMed  Tanaka N,  Fujita N,  Nishi A,  Satoh H,  Hosaka Y,  Ugaki M, et al.  The structure of starch can be manipulated by changing the expression levels of starch branching enzyme IIb in rice endosperm,  Plant Biotechnol. J. ,  2004, vol.  2 (pg.  507- 516) Google Scholar CrossRef Search ADS PubMed  Tateoka T.  On the systematic significance of starch grains of seeds in Poaceae,  J. Jap. Bot. ,  1954, vol.  29 (pg.  341- 347) Tateoka T.  Further studies on starch grains of seeds in Poaceae from the view point of systematics,  J. Jap. Bot. ,  1955, vol.  30 (pg.  199- 208) Tateoka T.  Starch grains of endosperm in grass systematics,  Bot. Mag. Tokyo ,  1962, vol.  75 (pg.  377- 383) Google Scholar CrossRef Search ADS   Temnykh S,  Park WD,  Ayres N,  Cartinhour S,  Hauck N,  Lipovich L, et al.  Mapping and genome organization of microsatellite sequences in rice (Oryza sativa L.),  Theor. Appl. Genet. ,  2000, vol.  100 (pg.  697- 712) Google Scholar CrossRef Search ADS   Walker JT,  Merritt NR.  Genetic control of abnormal starch granules and high amylose content in a mutant of Glacier barley,  Nature ,  1969, vol.  221 (pg.  482- 483) Google Scholar CrossRef Search ADS   Yano M,  Okuno K,  Kawakami J,  Satoh H,  Omura T.  High amylose mutants of rice, Oryza sativa L,  Theor. Appl. Genet. ,  1985, vol.  69 (pg.  253- 257) Google Scholar CrossRef Search ADS PubMed  Yun MS,  Kawagoe Y.  Amyloplast division progresses simultaneously at multiple sites in the endosperm of rice,  Plant Cell Physiol. ,  2009, vol.  50 (pg.  1617- 1626) Google Scholar CrossRef Search ADS PubMed  © The Author 2010. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Plant and Cell Physiology Oxford University Press

A Rapid, Direct Observation Method to Isolate Mutants with Defects in Starch Grain Morphology in Rice

Download PDF
14 pages

Loading next page...
 
/lp/oxford-university-press/a-rapid-direct-observation-method-to-isolate-mutants-with-defects-in-f6daAx7CpU

References (56)

Publisher
Oxford University Press
Copyright
© The Author 2010. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org
ISSN
0032-0781
eISSN
1471-9053
DOI
10.1093/pcp/pcq040
pmid
20360021
Publisher site
See Article on Publisher Site

Abstract

Abstract Starch forms transparent grains, referred to as starch grains (SGs), in amyloplasts. Despite the simple glucose polymer composition of starch, SGs exhibit different morphologies depending on plant species, especially in the endosperm of the Poaceae family. This study reports a novel method for preparing thin sections of endosperm without chemical fixation or resin embedding that allowed us to visualize subcellular SGs clearly. Using this method, we observed the SG morphologies of >5,000 mutagenized rice seeds and were able to isolate mutants in which SGs were morphologically altered. In five mutants, named ssg (substandard starch grain), increased numbers of small SGs (ssg1–ssg3), enlarged SGs (ssg4) and abnormal interior structures of SGs (ssg5) were observed. Amylopectin chain length distribution analysis and identification of the mutated gene suggested a possible allelic relationship between ssg1, ssg2, ssg3 and the previously isolated amylose-extender (ae) mutants, while ssg4 and ssg5 seemed to be novel mutants. Compared with conventional observation methods, the methods developed here are more effective for obtaining fine images of subcellular SGs and are suitable for the observation of a large number of samples. Introduction Starch is a biologically and commercially important polymer of glucose and is synthesized to form starch grains (SGs) inside plastids (amyloplasts) of higher plants (Buléon et al. 1998, Hancock and Tarbet 2000). Starch is water insoluble and hence osmotically inactive, making it a suitable long-term storage carbohydrate for seeds and tubes of many plant species. Starch contains two major components, amylose and amylopectin. Amylose is a linear molecule consisting of α-1,4-linked d-glucose chains, whereas amylopectin contains α-1,6-linked branches. Amylopectin is a major component of starch, accounting for 65–85% of its weight (w/w) (Nakamura 2002, James et al. 2003). Despite the simple composition of glucose polymers, SGs exhibit various morphologies depending on the plant species (Harz 1880, Tateoka 1954, Tateoka 1955, Tateoka 1962, Czaja 1978, Jane et al. 1994, James et al. 2003, Shapter et al. 2008). Although SG morphological diversity was reported >100 years ago, the underlying molecular mechanisms that account for the differences in morphology have not yet been determined. Morphologically, SGs are classified as either compound or simple grains (Harz 1880). Compound grains consist of many starch granules, and simple grains are composed of a single starch granule in an amyloplast. According to previous classification studies, Poaceae species are divided into four types based on the SG morphologies of their endosperm (Tateoka 1962). Compound grain types such as rice (Oryza sativa) possess only compound grains in their endosperm. Simple grain types are further divided into two subtypes referred to as bimodal and uniform. Bimodal types produce small and much larger simple SGs that co-exist in the same cells. Uniform types possess similar sized hexagonal, pentagonal or round, simple SGs. Barley (Hordeum vulgare) and wheat (Triticum aestivum) make bimodal SGs, and uniform SGs are found in maize (Zea mays). In addition to the above three types, a number of species, such as some belonging to the Andropogoneae tribe, show the co-existence of compound and simple grains inside the same cell (Tateoka 1955). Several starch-related mutants have been isolated in cereals by forward genetic approaches based on (i) the physical characteristics of seeds; (ii) the activities of starch biosynthesis enzymes; and (iii) the chemical properties of starch (Dvonch et al. 1951, Walker and Merritt 1969, Jarvi and Eslick 1975, Satoh and Omura 1981, Satoh et al. 2003a, Satoh et al. 2003b, Kang et al. 2005, Satoh et al. 2008). In addition, mutants defective in starch biosynthesis enzymes were isolated by reverse genetic approaches using rice Tos17 tagging systems (Fujita et al. 2007, Fujita et al. 2009). Some starch-related mutants exhibit SG morphologies different from those of wild-type strains (Dvonch et al. 1951, Yano et al. 1985, Burton et al. 2002, Fujita et al. 2003, Li et al. 2007); however, a molecular explanation for these morphological differences of SGs has not been proposed. Scanning electron microscopy (SEM) is normally used to observe SGs in extracted conditions where compound SGs are broken and the interior starch granules are detached from each other (Jane et al. 1994). To acquire images of subcellular SGs, cell walls need to be removed. Therefore, cells whose cell walls collapse during slicing of the endosperm make good subjects for SEM analysis (Shapter et al. 2008). Compared with SEM, light microscopic observation of endosperm thin sections is suitable for observing subcellular SGs, because they can be observed in most cells in sections. However, chemical fixation and resin embedding processes are needed for thin sectioning, which makes it unsuitable for large sample numbers, such as those found in a genetic screen. In this study, we describe a rapid method to prepare thin sections of cereal endosperm without chemical fixation or resin embedding. Using this method, we performed a genetic screen and isolated five rice mutants defective in SG morphologies. Three of the five mutants seemed to be allelic to the previously isolated amylose-extender (ae) mutants (Yano et al. 1985, Mizuno et al. 1993) and the other two appear to be novel mutants. Compared with conventional observation methods, the methods presented here are more effective for obtaining clear images of subcellular SGs and are highly suitable for the examination of a large number of samples. Results A rapid method to prepare thin sections of endosperm We developed a rapid method to prepare thin sections of mature endosperm (Fig. 1). A mature rice seed was fitted into a truncated 200 μl pipet tip (Fig. 1A). For rice and barley, truncated 200 μl tips were appropriate. For maize, a truncated 1 ml pipet tip was suitable. The seed-embedded tip was fixed on a block trimmer that was originally developed for resin block trimming for ultramicrotomy (Fig. 1B, C). The fixed seeds on the block trimmer were manipulated under a stereo microscope (Fig. 1D). The block trimmer was held by the third and fourth fingers of the non-dominant hand (Fig. 1E), while the index finger and thumb of the dominant hand held a razor blade attached to the seed. During trimming, the blade was kept horizontal to the seed (Fig. 1E, inset). The blade was also supported by the index finger of the non-dominant hand to adjust trimming pressure. Trimming generated a smooth surface on the top of the seed, which was exposed approximately 1 mm out of the pipet tip. Thin sectioning was performed in the same way as trimming with the same hand positions, but we kept the angle of the blade at approximately 30° (Fig. 1F). The seeds were easily trimmed using a razor blade (Fig. 1G, H), and thin sections were shaved off the endosperm (Fig. 1I, J). Forceps were used to take and place thin sections onto glass slides for staining (Fig. 1K). Fig. 1 View largeDownload slide The rapid method for preparation of thin sections of rice endosperm. (A) A rice seed was inserted into a truncated, disposable pipet tip (right). The inset is the magnified image of the captured seed. Bar = 5 mm. (B, C) The seed-captured tip was fixed on a block trimmer that was originally developed for resin block trimming for ultramicrotomy. (D) The block trimmer (indicated by a red arrow) can be manipulated under a stereo microscope. (E) Positions of the hands, fingers, a razor blade and the block trimmer equipped with the seed-captured tip for trimming. The inset is the magnified side view image of the captured seed and razor blade (indicated by an arrowhead). (F) A side view of the hand positions for thin sectioning. The angle of the blade is approximately 30° (inset, arrowhead). (G–K) Seed images under the stereo microscope. The razor blade and forceps are indicated by asterisks and a white arrow, respectively. Bars = 1 mm. (G) Before trimming. (H) The smooth surface after trimming. (I–J) A thin section shaved off the endosperm using the razor blade. (K) Forceps were used to take thin sections. Fig. 1 View largeDownload slide The rapid method for preparation of thin sections of rice endosperm. (A) A rice seed was inserted into a truncated, disposable pipet tip (right). The inset is the magnified image of the captured seed. Bar = 5 mm. (B, C) The seed-captured tip was fixed on a block trimmer that was originally developed for resin block trimming for ultramicrotomy. (D) The block trimmer (indicated by a red arrow) can be manipulated under a stereo microscope. (E) Positions of the hands, fingers, a razor blade and the block trimmer equipped with the seed-captured tip for trimming. The inset is the magnified side view image of the captured seed and razor blade (indicated by an arrowhead). (F) A side view of the hand positions for thin sectioning. The angle of the blade is approximately 30° (inset, arrowhead). (G–K) Seed images under the stereo microscope. The razor blade and forceps are indicated by asterisks and a white arrow, respectively. Bars = 1 mm. (G) Before trimming. (H) The smooth surface after trimming. (I–J) A thin section shaved off the endosperm using the razor blade. (K) Forceps were used to take thin sections. When the seed-embedded tip was attached to an ultramicrotome holder, thin sections could also be obtained using a glass knife and an ultramicrotome. Thin sections prepared by hand did not have a uniform thickness; however, the thinnest part of the thin sections prepared by hand was estimated to be approximately 1 μm thick, because 1 μm thin sections cut by the ultramicrotome produced a similar light transmission to the thinnest part of the hand-prepared thin sections (data not shown). To test the effectiveness of this method, we tried to prepare thin sections of three cereal species: rice, maize and barley (Fig. 2A–C). Thin sections prepared using the rapid method were stained with iodine and examined under a microscope. The morphologies of SGs were clearly observed (Fig. 2D–F). Rice SGs showed compound grains in which smaller granules were assembled (Fig. 2D, G). Occasionally, simple grains were also observed. Maize had simple grains that were round and uniform in size (Fig. 2E, H). In barley, we observed small (approximately 5 μm in diameter) and larger (approximately 20 μm in diameter) simple grains co-existing in the same cell (Fig. 2F, I). These SG morphologies were consistent with previous reports (Tateoka 1962) and SEM images (Fig. 2J–L). Fig. 2 View largeDownload slide Starch grain (SG) morphologies of three cereal species. (A–C) Seed images of rice (Oryza sativa), maize (Zea mays) and barley (Hordeum vulgare). Bars = 1 mm. (D–F) Iodine-stained sections of endosperm cells prepared using the rapid method. Bars = 20 μm. Boxes indicated by white dotted lines are illustrated areas in G–I. (G–I) Typical morphologies of SGs are illustrated. The SGs of rice are compound grains composed of many smaller granules. Maize SGs are the uniform type that are similar in size and round shaped. Barley SG is a bimodal type containing small and larger grains co-existing in the same cell. Red lines indicate cell walls. (J–L) SEM images. Bars = 20 μm. Fig. 2 View largeDownload slide Starch grain (SG) morphologies of three cereal species. (A–C) Seed images of rice (Oryza sativa), maize (Zea mays) and barley (Hordeum vulgare). Bars = 1 mm. (D–F) Iodine-stained sections of endosperm cells prepared using the rapid method. Bars = 20 μm. Boxes indicated by white dotted lines are illustrated areas in G–I. (G–I) Typical morphologies of SGs are illustrated. The SGs of rice are compound grains composed of many smaller granules. Maize SGs are the uniform type that are similar in size and round shaped. Barley SG is a bimodal type containing small and larger grains co-existing in the same cell. Red lines indicate cell walls. (J–L) SEM images. Bars = 20 μm. Screening of the mutants with distinct starch grain morphologies To isolate morphologically distinct SG mutants, we screened ethylmethane sulfonate (EMS)-treated M2 seeds from the lax panicle1 (lax1) mutant with a Nipponbare background using the rapid method. Because lax1 is a recessive mutation that results in the absence of lateral spikelets on panicles, sexual contamination could be monitored (Komatsu et al. 2001). We examined at least five seeds from each M2 line. Of the 1,152 M2 lines, five independent mutants with abnormal SG morphologies were isolated. All mutant grains showed a chalky endosperm (Fig. 3A). Subcellular images obtained by the rapid method showed that the morphologies of the parental lax1 and Nipponbare wild-type SGs were indistinguishable (Fig. 3B, Nipponbare lax1). This indicated that the lax1 mutation does not affect SG morphology. Normal SGs were uniform in size and approximately 10–20 μm in diameter. In ssg1, ssg2 and ssg3 mutants, many smaller, simple grains (<10 μm in diameter) were observed, in addition to the normal compound grains (Fig. 3B). In ssg4 mutants, larger SGs (>30 μm in diameter) were detected. SGs without internal compound structures were observed in ssg5 mutants. All of these phenotypes were inherited in subsequent generations. Fig. 3 View largeDownload slide Phenotypes of the isolated mutants. (A) Whole seeds (upper panel) and their cross-sections (lower panel) of the parental line (Nipponbare lax1) and mutants. Bars = 1 mm. (B) Iodine-stained endosperm sections prepared using the rapid method. Bars = 20 μm. Fig. 3 View largeDownload slide Phenotypes of the isolated mutants. (A) Whole seeds (upper panel) and their cross-sections (lower panel) of the parental line (Nipponbare lax1) and mutants. Bars = 1 mm. (B) Iodine-stained endosperm sections prepared using the rapid method. Bars = 20 μm. Starch grain morphologies of the ssg mutants observed using a conventional resin embedding method The rapid method described above allowed us to visualize SGs rapidly and easily. However, fine, thin sections were difficult to obtain from mutants with a floury endosperm, such as ssg4 and ssg5. To obtain fine images of SGs, we chemically fixed endosperm samples and embedded them in Technovit resin. Technovit sections (approximately 1 μm thick) were prepared and stained with iodine (Fig. 4). For the wild-type Nipponbare, the SG morphologies in the central and peripheral parts of the endosperm were similar to those obtained by the rapid method (Fig. 4A–D). In the central part of the ssg1 endosperm, most cells contained compound and small simple grains simultaneously (Fig. 4E, F), whereas the peripheral endosperm cells below the aleurone layer were filled with only compound grains (Fig. 4G, H). Fig. 4 View largeDownload slide Iodine-stained Technovit sections from ssg mutant endosperm. (A–D) Nipponbare, (E–H) ssg1, (I, J) ssg2, (K, L) ssg3, (M–P) ssg4, (Q–X) ssg5. The images shown in C, D, G, H, O, P, W and X are peripheral endosperm cells below the aleurone layer. Other images are from the central part of the endosperm. Asterisks indicate aleurone cells. In F, J and L, arrowheads and arrows indicate the positions of some small simple grains and compound grains, respectively. Bars = 20 μm. Fig. 4 View largeDownload slide Iodine-stained Technovit sections from ssg mutant endosperm. (A–D) Nipponbare, (E–H) ssg1, (I, J) ssg2, (K, L) ssg3, (M–P) ssg4, (Q–X) ssg5. The images shown in C, D, G, H, O, P, W and X are peripheral endosperm cells below the aleurone layer. Other images are from the central part of the endosperm. Asterisks indicate aleurone cells. In F, J and L, arrowheads and arrows indicate the positions of some small simple grains and compound grains, respectively. Bars = 20 μm. In the ssg2 and ssg3 mutants, the SG morphologies were quite similar to that of ssg1 (Fig. 4I–L). As in ssg1 mutants, the peripheral endosperm cells below the aleurone layer in ssg2 and ssg3 had only compound grains (data not shown). ssg4 SGs were so large that only a few SGs filled a single cell in the central part of the endosperm (Fig. 4M, N). This was also observed in the peripheral endosperm cells (Fig. 4O, P). In ssg5, SG morphologies varied between cells (Fig. 4Q–X). Some SGs stained weakly with iodine and their internal compound structures were not detected (Fig. 4R, S). Other SGs stained well with iodine, but their interior compound structures were abolished, especially in the central parts of the SGs (Fig. 4T–V). Abnormal SGs were most abundant in the central parts of endosperm cells, while normal-looking SGs were enriched in the peripheral endosperm cells below the aleurone layer (Fig. 4W, X). The structure of endosperm amylopectin in the mutant lines The chain length distribution of amylopectin is one of the molecular properties that determine amylopectin structure. Previously identified starch-related mutants often showed characteristic chain length distributions (Nakamura 2002). To evaluate the effect of ssg mutations on amylopectin structure, the chain length distributions [degree of polymerization (DP) ≤80] for isoamylolysate of endosperm amylopectin were analyzed using capillary electrophoresis (Fig. 5A–E). In ssg1 mutants, the proportion of short chains with DP ≤13 was markedly reduced, whereas the proportion of long chains with DP ≥38 was elevated (Fig. 5A). In particular, the intermediate length chains (14 ≤ DP ≤37) were significantly higher compared with the wild type. The same trend was observed in the ssg2 and ssg3 mutants (Fig. 5B, C). Compared with the wild type, no significant changes were detected in the ssg4 and ssg5 mutants (Fig. 5D, E). Fig. 5 View largeDownload slide Amylopectin structures and the accumulation of starch-branching enzymes in ssg mutants. (A–E) Chain length distributions of endosperm amylopectin from the parental line (black bars) and the ssg1, ssg2, ssg3, ssg4 and ssg5 mutants (represented by red, blue, green, purple and brown bars, respectively). The inset graphs compare the chain length distribution patterns (Δ molar %) between each mutant and the wild type. Values for the molar % and Δ molar % are the averages for three grains arbitrarily chosen from a single homozygous plant. Relative SDs of the molar % for each chain length from DP 4 to 80 were <0.3%. DP, degree of polymerization. (F) Immunoblot analysis using anti-BEIIb and anti-BEI antibodies. The proteins were extracted from the three mature seeds used for the chain length distribution analysis. Fig. 5 View largeDownload slide Amylopectin structures and the accumulation of starch-branching enzymes in ssg mutants. (A–E) Chain length distributions of endosperm amylopectin from the parental line (black bars) and the ssg1, ssg2, ssg3, ssg4 and ssg5 mutants (represented by red, blue, green, purple and brown bars, respectively). The inset graphs compare the chain length distribution patterns (Δ molar %) between each mutant and the wild type. Values for the molar % and Δ molar % are the averages for three grains arbitrarily chosen from a single homozygous plant. Relative SDs of the molar % for each chain length from DP 4 to 80 were <0.3%. DP, degree of polymerization. (F) Immunoblot analysis using anti-BEIIb and anti-BEI antibodies. The proteins were extracted from the three mature seeds used for the chain length distribution analysis. The chain length distribution patterns of ssg1, ssg2 and ssg3 amylopectin were quite similar to that of the previously identified ae mutant (Nishi et al. 2001, Nakamura 2002). In ae mutants, one of the starch-branching enzymes, BEIIb, is down-regulated (Mizuno et al. 1993). Starch-branching enzymes generate α-1,6 linkages by cleaving internal α-1,4 bonds and transferring the released, reducing ends to a C6 hydroxyl. Lack of BEIIb activity in ae results in abnormal amylopectin structures, which are enriched in long chains and depleted of short chains (Nishi et al. 2001). To investigate the accumulation of BEIIb in ssg mutants, total proteins were extracted from the mutant seeds and subjected to immunoblot analysis using anti-BEIIb antibodies (Fig. 5F). ssg1 seeds totally lacked BEIIb, whereas the other mutant seeds accumulated BEIIb to levels similar to the wild type. In contrast, the other starch-branching enzyme, BEI (Nakamura and Yamanouchi 1992), was expressed normally in all mutants. This indicates that, similar to the ae mutant, the abnormal amylopectin structure in ssg1 is caused by the lack of BEIIb. Genetic lesions in the BEIIb gene in the ssg1, ssg2 and ssg3 mutants When ssg1 was crossed with Kasalath, 20 out of 109 F2 seeds showed the co-existence of simple and compound grains in their endosperm, indicating that ssg1 segregated as a single recessive allele (χ2 = 2.6, P = 0.11). ssg1 was mapped to the middle of chromosome 2 where the BEIIb gene (Os02g0528200) is located (Harrington et al. 1997). Two splicing variants of the BEIIb gene were predicted based on expressed sequence tag (EST) analysis (LOC_Os02g32660.1 and LOC_Os02g32660.2, http://rice.plantbiology.msu.edu/). The shorter spliced variant is truncated relative to the longer transcript (Fig. 6A). We determined the genomic sequence of the longer BEIIb gene from the 5′ untranslated region (UTR) through to the 3′ UTR in the ssg1, ssg2 and ssg3 mutants, and found several base changes (Fig. 6A). In ssg1, two base changes were identified in introns. One was located in the 12th intron splicing acceptor site. The guanine residue (+5789) essential for correct mRNA splicing was replaced by adenine. The ssg2 mutant also had two base changes. One was located in the 18th exon, which caused an amino acid substitution from proline to leucine. In ssg3, a base change was found in the 17th exon, resulting in a glycine to arginine substitution. Fig. 6 View largeDownload slide Mutation sites in the BEIIb gene from the ssg1, ssg2 and ssg3 mutants. (A) Structure of the BEIIb gene (Os02g0528200). Two distinct splicing variants were predicted based on the EST analysis (LOC_Os02g32660.1, upper; LOC_Os02g32660.2, lower). The coding and untranslated regions are depicted as blue and white boxes, respectively. Introns are indicated by black bars. Adenine from the translation start codon (ATG) is designated as +1, and the stop codons (TGA) are +11,025 and +8,428, respectively. The ssg1 mutant had two base pair changes (G to A) at +5,789 and +10,703. The former change was located at the 5′ splicing acceptor site of the 12th intron. The ssg2 mutant also had two base pair changes (T to C and C to T) at +7,863 and +9,109, respectively. The latter change was located in the 18th exon and resulted in an amino acid substitution of Pro644 by leucine (Leu). The ssg3 mutant had a single base pair change (G to A) in the 17th exon, leading to an amino acid substitution of Gly667 by arginine (Arg). (B) The protein structure of the Oryza sativa BEIIb protein. Dark and light gray boxes represent the transit peptide and mature regions, respectively. Black boxes are the four catalytic sites conserved in amylolytic enzymes (region I, II, III and IV). The predicted positions of the β-strands and α-helices of the (β/α)8 barrel domain common to the amylolytic superfamily proteins are indicated with white and red boxes, respectively. Green and blue arrowheads indicate the mutation sites in ssg3 and ssg2, respectively. The open box is the region that was used for the alignment in C. (C) Alignment with other starch- and glycogen-branching enzymes. Enzymes indicated with blue and red characters belong to families A and B of the starch-branching enzymes, respectively. Glycogen-branching enzymes are indicated in black. OsBEIIb and OsBEI (Oryza sativa, 02g0528200 and Os06g0726400), ZmBEIIb and ZmBEI (Zea mays, NP_001105316 and NP_001105370), TaBEII and TaBEI (Triticum aestivum, Y11282 and Y12320), PsSBEII and PsSBEI (Pisum sativum, CAA56320 and CAA56319), StSBEII and StSBEI (Solanum tuberosum, CAB40748 and CAA49463), ScGLC3 (Saccharomyces cerevisiae, AAA34632) and HsGBEI (Homo sapiens, NM_000158). Perfectly conserved residues are shown in black. Blue and green arrowheads indicate the residues substituted by ssg3 and ssg2 mutations, respectively. The Clustal X program (Larkin et al. 2007) was used for the alignment. Fig. 6 View largeDownload slide Mutation sites in the BEIIb gene from the ssg1, ssg2 and ssg3 mutants. (A) Structure of the BEIIb gene (Os02g0528200). Two distinct splicing variants were predicted based on the EST analysis (LOC_Os02g32660.1, upper; LOC_Os02g32660.2, lower). The coding and untranslated regions are depicted as blue and white boxes, respectively. Introns are indicated by black bars. Adenine from the translation start codon (ATG) is designated as +1, and the stop codons (TGA) are +11,025 and +8,428, respectively. The ssg1 mutant had two base pair changes (G to A) at +5,789 and +10,703. The former change was located at the 5′ splicing acceptor site of the 12th intron. The ssg2 mutant also had two base pair changes (T to C and C to T) at +7,863 and +9,109, respectively. The latter change was located in the 18th exon and resulted in an amino acid substitution of Pro644 by leucine (Leu). The ssg3 mutant had a single base pair change (G to A) in the 17th exon, leading to an amino acid substitution of Gly667 by arginine (Arg). (B) The protein structure of the Oryza sativa BEIIb protein. Dark and light gray boxes represent the transit peptide and mature regions, respectively. Black boxes are the four catalytic sites conserved in amylolytic enzymes (region I, II, III and IV). The predicted positions of the β-strands and α-helices of the (β/α)8 barrel domain common to the amylolytic superfamily proteins are indicated with white and red boxes, respectively. Green and blue arrowheads indicate the mutation sites in ssg3 and ssg2, respectively. The open box is the region that was used for the alignment in C. (C) Alignment with other starch- and glycogen-branching enzymes. Enzymes indicated with blue and red characters belong to families A and B of the starch-branching enzymes, respectively. Glycogen-branching enzymes are indicated in black. OsBEIIb and OsBEI (Oryza sativa, 02g0528200 and Os06g0726400), ZmBEIIb and ZmBEI (Zea mays, NP_001105316 and NP_001105370), TaBEII and TaBEI (Triticum aestivum, Y11282 and Y12320), PsSBEII and PsSBEI (Pisum sativum, CAA56320 and CAA56319), StSBEII and StSBEI (Solanum tuberosum, CAB40748 and CAA49463), ScGLC3 (Saccharomyces cerevisiae, AAA34632) and HsGBEI (Homo sapiens, NM_000158). Perfectly conserved residues are shown in black. Blue and green arrowheads indicate the residues substituted by ssg3 and ssg2 mutations, respectively. The Clustal X program (Larkin et al. 2007) was used for the alignment. To assess the linkage between the base change and the ssg3 SG phenotype, we next designed the derived cleaved-amplified polymorphic sequence (dCAPS) primers according to the base change in ssg3 (+8,898 in Fig. 6A). The dCAPS successfully detected the genotypes of wild-type, ssg3 and F1 heterozygous plants (Supplementary Fig. S1). When ssg3 was crossed with Kasalath, 21 out of 79 F2 seeds showed the SG phenotype specific to ssg3, indicating that ssg3 is single recessive (χ2 = 0.11, P = 0.75). Genotyping of the 21 mutant seeds using the dCAPS primer showed that all 21 seeds were homozygous for the ssg3 base change (Supplementary Fig. S1). This result supports the idea that the base change in the BEIIb gene in the ssg3 mutant is responsible for the ssg3 SG phenotype. Based on the structural analysis of other glucanases, branching enzymes are predicted to contain a catalytic (β/α)8 barrel domain that acts as a scaffold for substrate binding and catalysis (Buisson et al. 1987, Jespersen et al. 1993, Abad et al. 2002). The substituted amino acids in ssg2 and ssg3 were located inside the loop between the eighth β-strand and the α-helix, and outside the conserved catalytic domain (regions I, II, III and IV, Fig. 6B) (Svensson 1994). Plant starch-branching enzymes are separated into two families, A and B, based on their amino acid sequence similarities (Burton et al. 1995, Mizuno et al. 2001). Rice BEIIb belongs to family A and BEI is a member of family B. The proline residue substituted by the ssg2 mutation is conserved in both families, while the glycine residue substituted in ssg3 is conserved only in family A (Fig. 6C). Discussion Usefulness of the rapid method for preparing thin sections of endosperm It has been well documented that SG morphologies show significant differences between plant species (Harz 1880, Tateoka 1954, Tateoka 1955, Tateoka 1962, Czaja 1978, Jane et al. 1994, James et al. 2003, Shapter et al. 2008). Various morphologies of SGs in many plant species have been observed in extracted samples; however, subcellular SGs have rarely been observed except for those found in major cereal species. In this study, we developed a method to prepare thin sections of cereal endosperm (Fig. 1) that allowed us to visualize the subcellular SGs in endosperm cells rapidly and easily (Fig. 2). Another advantage of this method was that seeds used for sectioning remained alive and ready to grow because chemical fixation was not required. This enabled easier genetic screening based on the direct observation of SGs (Fig. 3). Using this method, we found that >200 rice seeds could be examined in a single day (data not shown). In addition, for wild-type seeds, the images obtained using the rapid method were quite similar to those obtained by conventional resin sectioning methods (Figs. 3B, 4A–D). However, in seeds with floury properties, such as those from ssg4 and ssg5, fine, thin sections were hard to prepare using the rapid method. Therefore, the images of sections obtained by the rapid method were lower in quality than for sections obtained using resin sectioning methods (Figs. 3B, 4M–X). Possible allelic relationship between ssg1, ssg2, ssg3 and ae mutants Out of five ssg mutants isolated in this study, three (ssg1, ssg2 and ssg3) had mutations in the BEIIb gene (Fig. 6). Rice ae mutants were originally isolated because of their increased amylose content (Yano et al. 1985). Further analysis revealed that the amylopectin structure, not amylose levels, is affected in ae, due to the lack of the BEIIb protein (Mizuno et al. 1993, Nishi et al. 2001). The mutation sites of rice ae have not yet been determined. However, the genomic region covering the BEIIb gene can complement the altered structure of ae amylopectin, suggesting that defective BEIIb is responsible for ae (Tanaka et al. 2004). ae SGs observed by SEM display irregular, round-shaped morphologies (Yano et al. 1985, Tanaka et al. 2004). The number of round-shaped simple grains increased dramatically in ssg1, ssg2 and ssg3 (Fig. 4E, F, I–L). In addition, the amylopectin chain length distributions (Fig. 5) were very similar in ssg1, ssg2, ssg3 and ae mutants. Together, these data suggest a possible allelic relationship between ssg1, ssg2, ssg3 and ae. In ssg1, the point mutation (+5,789) was located at a guanine residue at the splicing acceptor site (Fig. 6A). Thus, a lack of mature mRNAs probably leads to a significant reduction in BEIIb protein levels (Fig. 5F). In ssg2 and ssg3, point mutations (+9,109 and +8,898, respectively) are expected to cause amino acid substitutions; however, these mutations had no effect on the accumulation of the BEIIb protein (Fig. 5F). Previous domain-swapping experiments between maize BEI and BEII showed that the C-terminal domain of BE is involved in substrate specificity (Kuriki et al. 1997). ssg2 and ssg3 mutations were located in the C-terminal domain used in the domain-swapping experiments. Collectively, these data suggest that the amino acids substituted in ssg2 and ssg3 are crucial for BEIIb activity, and that the accumulated BEIIb proteins in ssg2 and ssg3 are enzymatically inactive. ssg4 and ssg5 will provide new insights into amyloplast division and starch grain morphologies Genetic mapping of the ssg4 and ssg5 loci is now underway. Therefore, we have not yet determined whether they are novel mutants. Out of the previously isolated starch-related mutants, sugary-1 and floury 2 (flo2) mutants have some common characteristics with ssg4 and ssg5. ssg4 and ssg5 seeds showed floury phenotypes (data not shown) and some cells in ssg5 were stained weakly with iodine (Fig. 4R), which is characteristic of sugary-1 mutants (Nakamura et al. 1996). sugary-1 lacks one of the starch-debranching enzymes, isoamylase 1, which leads to the production of highly and randomly branched polyglucan (phytoglycogen) instead of amylopectin (Nakamura et al. 1997). Technovit sections from sugary-1 endosperm showed that some SGs in the central part of the endosperm were stained weakly by iodine and were strikingly amorphous (Supplementary Fig. S2A–C). Consistent with previous observations (Kawagoe et al. 2005), compound grains consisting of much smaller starch granules were also detected (Supplementalry Fig. S2D). In addition to the abnormal SGs, normal SGs were also observed in the central and peripheral parts of the sugary-1 endosperm (Supplementary Fig. S2E–H). In the case of flo2, SGs showed normal morphologies and sizes (Supplementary Fig. S2I, J), indicating that ssg4 and ssg5 are different from flo2. The morphological variations of SGs and the presence of cells that stain weakly with iodine are commonly observed in ssg5 and sugary-1 (Fig. 4Q–S and Supplementary Fig. S2A–C); however, they appear to be different mutants for the following reasons. In sugary-1, the phytoglycogen-concentrated region that is not stained with iodine dominates a large portion of the central part of the endosperm (Nakamura et al. 1997), whereas the weakly stained region was not obvious in ssg5 under a stereo microscope (Supplementary Fig. S3). Furthermore, the SGs that contained a simple globular structure in the central part, as shown in Fig. 4 (Q and U), were specific to ssg5. The chain length distribution of amylopectin in ssg5 and the wild type was quite similar (Fig. 5E), suggesting that the gene responsible for ssg5 is not related to amylopectin biosynthesis. In contrast, sugary-1 amylopectin showed significant increases in the number of short chains (DP ≤10), and a depletion of intermediate (10≤ DP ≤30) and long chains (40≤ DP), because of a lack of the starch-debranching enzyme, isoamylase 1 (Kubo et al. 1999). Therefore, we concluded that ssg5 is a different mutation from sugary-1 and may be a novel factor required to determine or maintain SG morphologies. The ssg4 SGs were an enlarged size, while the cell size was significantly smaller compared with the wild type (Fig. 4A–D, M–P). Recent studies using Arabidopsis mutants have demonstrated that inhibition of chloroplast division causes chloroplast enlargement (Glynn et al. 2007). For example, the Arabidopsis mutant arc5 (accumulation and replication of chloroplast 5) has only three chloroplasts in a single cell (Gao et al. 2003). ARC5 encodes a member of the dynamin superfamily, which plays a central role in chloroplast division (Gao et al. 2003). Similarly, the division of amyloplasts may be inhibited in ssg4 endosperm. Rice arc5 mutants have been isolated and reported to have enlarged chloroplasts (Yun and Kawagoe 2009). However, rice arc5 mutants develop pleomorphic amyloplasts that are sometimes elongated and have a beads-on-a-string structure (Yun and Kawagoe 2009). The morphologies of ssg4 and arc5 amyloplasts are different from each other, suggesting that ssg4 is distinct from arc5. Different division mechanisms between chloroplasts and amyloplasts have been proposed (Yun and Kawagoe 2009). In leaves, chloroplasts divide by binary fission, in which fission of the thylakoid membrane precedes the separation of the envelope membrane. During binary fission, dividing chloroplasts often adopt dumbbell-like shapes (Glynn et al. 2007). In amyloplasts, two different division mechanisms have been suggested: (i) the multiple, beads-on-a-string-type; and (ii) the budding-type. In multiple, beads-on-a-string-type division, elongated amyloplasts generate multiple constriction sites to which ARC5 and another chloroplast division-related factor, FtsZ, are recruited. The beads-on-a-string structure is observed when the endosperm cell is still small and rapidly expanding in volume. At later stages of endosperm development, large amyloplasts divide by the protrusion of smaller amyloplasts from their surfaces. This protrusion process is called budding-type division and is achieved by unknown mechanisms. SSG4 may encode a novel factor that has not been discovered by previous, chloroplast-focused studies, and that may be involved in amyloplast budding-type division. Morphological variations of starch grains between peripheral and central parts of the endosperm SG morphologies affected by ssg mutations varied between peripheral and central parts of the endosperm (Fig. 4). In ssg mutants, except for ssg4, most SGs in the peripheral endosperm cells looked normal, while abnormal SGs were enriched in the central portion of the endosperm. This pattern was also observed in the sugary-1 mutant (Supplementary Fig. S2A–H), suggesting the general character of SG morphologies. In rice, SG formation begins in the innermost cells of the endosperm and spreads to outer cells centrifugally. The peripheral endosperm cells are the last parts to be filled with SGs and this process is completed about 1 month after the start of SG formation in the central region (Hoshikawa 1993). Gene expression of starch-related enzymes changes dramatically between the early and late stages of seed development (Ohdan et al. 2005). Therefore, the molecular mechanisms responsible for SG formation may also change within this time period. Detailed microscopic observation of SGs during all stages of endosperm development in ssg mutants is necessary for a better understanding of SG formation. The endosperm of all ssg mutants showed chalkiness that was limited to the central portions (Fig. 3A). Generally, the chalky parts were not sufficiently filled with SGs and had many minute gaps that caused light to scatter within these regions, giving them a white appearance (Hoshikawa 1993). Abnormal SG morphologies in ssg mutants may lead to cells that were incompletely filled with SGs and cause the chalky appearance, especially in the central part of the endosperm. However, in rare cases of ssg1, non-chalky, translucent grains were also obtained (Supplementary Fig. S4A, B) where SG morphologies were indistinguishable from those in the chalky ssg1 endosperm (Supplementary Fig. S4C). This suggests that the abnormal SG morphologies are not the direct cause of chalkiness in ssg1. Concluding remarks Many Poaceae species have been examined, mostly in extracted conditions, for SG morphologies (Tateoka 1962). We are now re-evaluating SG morphologies of the Poaceae species using the rapid and resin-embedding methods to verify previous observations and discover novel types of SGs with unique morphologies. Molecular analysis focusing on the mechanisms that determine SG morphology and descriptive analysis of cross-species SG diversity will lead to a more complete understanding of the molecular diversity of SG morphologies. The methods reported in this study should serve as an effective technique for these analyses. Materials and Methods Plant material and growth The endosperm mutant lines used in this experiment were selected from lax1 mutants that were originally obtained from regenerated plants derived from Nipponbare calli. Mutagenization was carried out by soaking lax1 seeds in 1.5% (v/v) methanesulfonic acid ethyl ester (Sigma, Tokyo, Japan). The M2 line derived from a single M1 plant were grown and M2 seeds were collected from individual M1 plants after self-fertilization. Rice plants were grown at an experimental field of the Institute of Plant Science and Resources, Okayama University under natural conditions or at 28°C in a greenhouse. Seeds of Zea mays (maize) were purchased from Nagano-ken Yuukiseisansharengou Ltd. (Nagano, Japan). Seeds of the sugary-1 mutant (Tankei 2013) were provided by the Genebank of the National Institute of Agrobiological Sciences. floury 2 mutants were previously isolated from the γ-ray-treated M2 population (Maekawa 1985). Isolation of starch grain mutants by the rapid observation method Screening was carried out with at least five seeds from each M2 line (1,152 lines). Endosperm thin sections from the M2 seeds were prepared by the rapid method. A mature M2 seed was inserted into a truncated pipet tip (Fig. 1A, 200 μl pipet tip, #110, Quality Scientific Plastics, Petaluma, CA, USA). The seed-captured tip was fixed on the block trimmer (Fig. 1B, C; Okenshoji, Tokyo, Japan) that was originally developed for resin block trimming for ultramicrotomy. The fixed seeds on the block trimmer were manipulated under a stereo microscope (Fig. 1D, SZ61, Olympus, Tokyo, Japan) and were trimmed with a razor blade (#FH-10, Feather anzen kamisori, Osaka, Japan) to generate a smooth surface on the top of the seed. The positions of the hands, fingers, a razor blade and the block trimmer during trimming and thin sectioning are described in the Results section. Starch staining was done by immersion in a drop of deionized water containing 40 times diluted Lugol solution (iodine/potassium iodine solution, MP Biomedicals, Eschwege, Germany). The samples were subsequently examined under a microscope (AX70, Olympus). Scanning electron microscopy To observe endosperm starch grains by SEM, dry seeds of rice, maize and barley were cut with a razor blade to expose the fractured surfaces of the endosperm. The specimens were coated with gold by a sputter coating machine (MSP-1S, Shinkuu device, Ibaraki, Japan) and examined with SEM (Quanta 250, FEI, Hillsboro, OR, USA). Thin sections of Technovit 7100 resin-embedded endosperm Approximately 1 mm cubic blocks were cut out from the endosperm of dry seeds and fixed in FAA solution containing 5% (v/v) formalin, 5% (v/v) acetic acid and 50% (v/v) ethanol, for at least 12 h at room temperature. Samples were subsequently dehydrated through a graded ethanol series [30, 50, 70, 90 and 100% (v/v)] and then embedded in Technovit 7100 resin (Kulzer and Company, Wehrheim, Germany). The embedded samples were cut with an Ultracut N ultramicrotome (Reichert-Nissei, Tokyo, Japan) and glass knives, and dried on coverslips. Thin sections (approximately 1 μm thickness) were stained with 40 times diluted Lugol solution in deionized water for at least 5 s and subsequently examined under a microscope (AX70). Chain length distribution of endosperm amylopectin by capillary electrophoresis To extract starch from mature endosperm for amylopectin chain length distribution, embryo removed from mature dry seeds was crushed with pliers and hand homogenized using a mortar and pestle. It was then suspended in 5 ml of methanol and boiled for 10 min. The homogenate was centrifuged at 2,500 × g for 10 min. The precipitate was washed twice with 5 ml of 90% (v/v) methanol and suspended in 300 μl of 0.25 M NaOH. The suspension was boiled for 5 min. A 9.6 μl aliquot of 100% acetic acid, 100 μl of 2% (w/v) NaN3 and 1,090 μl of distilled water were added to the gelatinized α-polyglucan sample. The sample was hydrolyzed by adding 4 μl of Pseudomonas amyloderamosa isoamylase (354 U; Hayashibara, Okayama, Japan) at 37°C for 24 h. The hydrolyzed sample was boiled for 20 min and centrifuged. The supernatant was deionized by filtration on an ion exchange resin [BioRad AG501-X8(D)] in a microtube. An appropriate aliquot containing 10 nmol of reduced end estimated by the modified Park Johnson method (Hizukuri et al. 1981) was evaporated to dryness in a centrifugal vacuum evaporator (Taitec, Tokyo, JApan). Fluorescence labeling and capillary electrophoresis were performed according to the previously reported method (O’Shea and Morell 1996) and the protocols provided by the manufacturer by using the eCAP N-linked oligosaccharide profiling kit and capillary electrophoresis (P/ACE MDQ Capillary Electrophoresis System; Beckman Coulter’s, Fullerton, CA, USA). Immunoblot analysis The rice flours were prepared from each mutant and the parental line and were mixed in 50 vols. of extraction buffer: 8 M urea, 5% (v/v) 2-mercaptoethanol, 4% (w/v) SDS, 125 mM Tris–HCl, pH 6.8. The samples were incubated at 37°C for 12 h and subjected to centrifugation at 20,000 × g at 20°C for 10 min. The 10 μl supernatant was mixed with 5 μl of SDS solution containing 55 mM Tris–HCl (pH 6.8), 2.3% SDS, 5% 2-mercaptoethanol and 10% glycerol, and subjected to 7.5% SDS–PAGE. After SDS–PAGE, the proteins were transferred electrophoretically to a polyvinylidene difuoride (PVDF) membrane (Millipore, Tokyo, Japan) by transblotter (Nihon Eido Co., Tokyo, Japan). The membrane was then incubated in Tris-buffered saline (pH 7.5) plus 0.05% (v/v) Tween-20 with the anti-BEIIb and anti-BEI antibodies (Nakamura et al. 1992) for 1 h. Dilutions of the antibodies are 1 : 1,000 (v/v). Horseradish peroxidase-conjugated goat antibodies against rabbit IgG (Bio-Rad, Hercules, CA, USA) were diluted (1 : 2,000) to be used as second antibodies. The immunoreactive bands were finally detected with 4-chloro-1-naphthol. Mapping of the SSG1 locus and determination of the mutation site of ssg1, ssg2 and ssg3 mutants For mapping of the SSG1 gene, we constructed an F2 population derived from a cross between the ssg1 mutant and Kasalath. Endosperm thin sections of each F2 seed were examined by the rapid method to select ssg1 mutant seeds. Rough mapping of this gene was done using 24 F2ssg1 plants. The genomic DNA of these ssg1 mutants was individually isolated and analyzed using simple sequence length polymorphism markers (Temnykh et al. 2000, McCouch et al. 2002, Maekawa et al. 2005) to determine the molecular markers linked to SSG1. SSG1 was defined to the region where BEIIb gene was located. BEIIb gene regions covering from the 5′ UTR to the 3′ UTR of ssg1, ssg2 and ssg3 were sequenced using the BigDye Terminators v1.1 cycle sequencing kit (Applied Biosystems, Foster City, CA, USA) and a 3100-Avant Genetic Analyzer (Applied Biosystems). Co-segregation analysis of the base change in the BEIIb gene with the ssg3 phenotype To confirm that the base change (+8,898 in Fig. 6A) in the ssg3 mutant co-segregates with the ssg phenotype, we crossed ssg3 with Kasalath and produced F2 seed populations. Out of the 79 F2 seeds, 21 ssg seeds were selected by the rapid method and genomes were isolated. To detect the base change in ssg3, dCAPS primers were designed as follows; 5′-TGAATTT TAATGAATACTGGCATGA-3′and 5′-AAAGTTAAGATAGCC TTCTCCTCCTGAGC-3′. The PCR conditions were as follows: 94°C for 2 min, and 35 cycles of 94°C for 30 s, 50°C for 45 s and 68°C for 1 min. The PCR product was digested with SacI and PCR products were subsequently separated by 15% PAGE and detected with ethidium bromide staining. In the case of ssg3, a PCR product (174 bp) was digested into 148 and 26 bp. In the case of the wild type (Nipponbare), the PCR product (174 bp) was not digested. Funding The Ministry of Education, Culture, Sports, Science and Technology [Grant-in-Aid for Scientific Research (No. 20770036 to R.M.)]; the Iijima Memorial Foundation for the Promotion of Food Science and Technology; the Science and Technology Foundation of Japan; the Oohara Foundation. Acknowledgments The authors would like to thank Dr. Shin Taketa (Institute of Plant Science and Resources, Okayama University) for providing the barley seeds, and Dr. Jun Yamashita (Institute of Plant Science and Resources, Okayama University) for critical reading of the manuscript and valuable suggestions. The seeds of sugary-1 mutant (Tankei 2013) were kindly provided by the Genebank of the National Institute of Agrobiological Sciences (Tsukuba, Japan). We would also like to thank Rie Hijiya and Rumiko Itoh for their technical assistance. Abbreviations Abbreviations ae amylose-extender BE branching enzyme dCAPS derived cleaved-amplified polymorphic sequence DP degree of polymerization EST expressed sequence tag SEM scanning electron microscopy SG starch grain ssg substandard starch grain UTR untranslated region. References Abad MC,  Binderup K,  Rios-Steiner J,  Arni RK,  Preiss J,  Geiger JH.  The X-ray crystallographic structure of Escherichia coli branching enzyme,  J. Biol. Chem. ,  2002, vol.  277 (pg.  42164- 42170) Google Scholar CrossRef Search ADS PubMed  Buisson G,  Duee E,  Haser R,  Payan F.  Three dimensional structure of porcine pancreatic α-amylase at 2.9 Å resolution. Role of calcium in structure and activity,  EMBO J. ,  1987, vol.  6 (pg.  3909- 3916) Google Scholar PubMed  Buléon A,  Colonna P,  Planchot V,  Ball S.  Starch granules: structure and biosynthesis,  Int. J. Biol. Macromol. ,  1998, vol.  23 (pg.  85- 112) Google Scholar CrossRef Search ADS PubMed  Burton RA,  Bewley JD,  Smith AM,  Bhattacharyya MK,  Tatge H,  Ring S, et al.  Starch branching enzymes belonging to distinct enzyme families are differentially expressed during pea embryo development,  Plant J. ,  1995, vol.  7 (pg.  3- 15) Google Scholar CrossRef Search ADS PubMed  Burton RA,  Jenner H,  Carrangis L,  Fahy B,  Fincher GB,  Hylton C, et al.  Starch granule initiation and growth are altered in barley mutants that lack isoamylase activity,  Plant J. ,  2002, vol.  31 (pg.  97- 112) Google Scholar CrossRef Search ADS PubMed  Czaja AT.  Structure of starch grains and the classification of vascular plant families,  Taxon ,  1978, vol.  27 (pg.  463- 470) Google Scholar CrossRef Search ADS   Dvonch W,  Kramer HH,  Whistler RL.  Polysaccharides of high-amylose corn,  Cereal Chem. ,  1951, vol.  28 (pg.  270- 280) Fujita N,  Kubo A,  Suh D-S,  Wong K-S,  Jane J-L,  Ozawa K, et al.  Antisense inhibition of isoamylase alters the structure of amylopectin and the physicochemical properties of starch in rice cndosperm,  Plant Cell Physiol. ,  2003, vol.  44 (pg.  607- 618) Google Scholar CrossRef Search ADS PubMed  Fujita N,  Toyosawa Y,  Utsumi Y,  Higuchi T,  Hanashiro I,  Ikegama A, et al.  Characterization of pullulanase (PUL)-deficient mutants of rice (Oryza sativa L.) and the function of PUL on starch biosynthesis in the developing rice endosperm,  J. Exp. Bot. ,  2009, vol.  60 (pg.  1009- 1023) Google Scholar CrossRef Search ADS PubMed  Fujita N,  Yoshida M,  Kondo T,  Saito K,  Utsumi Y,  Tokunaga T, et al.  Characterization of SSIIIa-deficient mutants of rice: the function of SSIIIa and pleiotropic effects by SSIIIa deficiency in the rice endosperm,  Plant Physiol. ,  2007, vol.  144 (pg.  2009- 2023) Google Scholar CrossRef Search ADS PubMed  Gao H,  Kadirjan-Kalbach D,  Froehlich JE,  Osteryoung KW.  ARC5, a cytosolic dynamin-like protein from plants, is part of the chloroplast division machinery,  Proc. Natl Acad. Sci. USA ,  2003, vol.  100 (pg.  4328- 4333) Google Scholar CrossRef Search ADS   Glynn JM,  Miyagishima S-Y,  Yoder DW,  Osteryoung KW,  Vitha S.  Chloroplast division,  Traffic ,  2007, vol.  8 (pg.  451- 461) Google Scholar CrossRef Search ADS PubMed  Hancock RD,  Tarbet BJ.  The other double helix—the fascinating chemistry of starch,  J. Chem. Educ. ,  2000, vol.  77 (pg.  988- 992) Google Scholar CrossRef Search ADS   Harrington SE,  Bligh HFJ,  Park WD,  Jones CA,  McCouch SR.  Linkage mapping of starch branching enzyme III in rice (Oryza sativa L.) and prediction of location of orthologous genes in other grasses,  Theor. Appl. Genet. ,  1997, vol.  94 (pg.  564- 568) Google Scholar CrossRef Search ADS   Harz CO.  Beiträge zur systematik der Gramineen,  Linnaea ,  1880, vol.  43 (pg.  1- 30) Hizukuri S,  Takeda Y,  Yasuda M,  Suzuki A.  Multi-branched nature of amylose and the action of debranching enzymes,  Carbohydr. Res. ,  1981, vol.  94 (pg.  205- 213) Google Scholar CrossRef Search ADS   Hoshikawa K.  Matsuo T,  Hoshikawa K.  Anthesis, fertilization and development of caryopsis,  Science of the Rice Plant. Vol. 1 Morphology ,  1993 Tokyo Food and Agriculture Policy Research Center(pg.  339- 376) James MG,  Denyer K,  Myers AM.  Starch synthesis in the cereal endosperm,  Curr. Opin. Plant Biol. ,  2003, vol.  6 (pg.  215- 222) Google Scholar CrossRef Search ADS PubMed  Jane J-L,  Kasemsuwan T,  Leas S,  Ames IA,  Zobel H,  Darien IL, et al.  Anthology of starch granule morphology by scanning electron microscopy,  Starch/Stärke ,  1994, vol.  46 (pg.  121- 129) Google Scholar CrossRef Search ADS   Jarvi AJ,  Eslick RF.  Shrunken endosperm mutants in barley,  Crop Sci. ,  1975, vol.  15 (pg.  363- 366) Google Scholar CrossRef Search ADS   Jespersen HM,  MacGregor EA,  Henrissat B,  Sierks MR,  Svensson B.  Starch- and glycogen-debranching and branching enzymes: prediction of structural features of the catalytic (β/α)8-barrel domain and evolutionary relationship to other amylolytic enzymes,  J. Protein Chem. ,  1993, vol.  12 (pg.  791- 805) Google Scholar CrossRef Search ADS PubMed  Kang HG,  Park S,  Matsuoka M,  An G.  White-core endosperm floury endosperm-4 in rice is generated by knockout mutations in the C4-type pyruvate orthophosphate dikinase gene (OsPPDKB),  Plant J. ,  2005, vol.  42 (pg.  901- 911) Google Scholar CrossRef Search ADS PubMed  Kawagoe Y,  Kubo A,  Satoh H,  Takaiwa F,  Nakamura Y.  Roles of isoamylase and ADP-glucose pyrophosphorylase in starch granule synthesis in rice endosperm,  Plant J. ,  2005, vol.  42 (pg.  164- 174) Google Scholar CrossRef Search ADS PubMed  Komatsu M,  Maekawa M,  Shimamoto K,  Kyozuka J.  The LAX1 and FRIZZY PANICLE 2 genes determine the inflorescence architecture of rice by controlling rachis-branch and spikelet development,  Dev. Biol. ,  2001, vol.  231 (pg.  364- 373) Google Scholar CrossRef Search ADS PubMed  Kubo A,  Fujita N,  Harada K,  Matsuda T,  Satoh H,  Nakamura Y.  The starch-debranching enzymes isoamylase and pullulanase are both involved in amylopectin biosynthesis in rice endosperm,  Plant Physiol. ,  1999, vol.  121 (pg.  399- 409) Google Scholar CrossRef Search ADS PubMed  Kuriki T,  Stewart DC,  Preiss J.  Construction of chimeric enzymes out of maize endosperm branching enzymes I and II: activity and properties,  J. Biol. Chem. ,  1997, vol.  272 (pg.  28999- 29004) Google Scholar CrossRef Search ADS PubMed  Larkin MA,  Blackshields G,  Brown NP,  Chenna R,  McGettigan PA,  McWilliam H, et al.  Clustal W and Clustal X version 2.0,  Bioinformatics ,  2007, vol.  23 (pg.  2947- 2948) Google Scholar CrossRef Search ADS PubMed  Li J-H,  Guiltinan MJ,  Thompson DB.  Mutation of the maize sbe1a and ae genes alters morphology and physical behavior of wx-type endosperm starch granules,  Carbohydr. Res. ,  2007, vol.  342 (pg.  2619- 2627) Google Scholar CrossRef Search ADS PubMed  Maekawa M.  Location of a floury endosperm gene in the second linkage group,  Rice Genet. Newslett. ,  1985, vol.  2 (pg.  57- 58) Maekawa M,  Takamure I,  Ahmed N,  Kyozuka J.  Bunketsu-waito, one of the tillering dwarf, is controlled by a single recessive gene in rice (Oryza sativa L.),  Breed. Sci. ,  2005, vol.  55 (pg.  193- 196) Google Scholar CrossRef Search ADS   McCouch SR,  Teytelman L,  Xu Y,  Lobos KB,  Clare K,  Walton M, et al.  Development and mapping of 2240 new SSR markers for rice (Oryza sativa L.),  DNA Res. ,  2002, vol.  9 (pg.  199- 207) Google Scholar CrossRef Search ADS PubMed  Mizuno K,  Kawasaki T,  Shimada H,  Satoh H,  Kobayashi E,  Okumura S, et al.  Alteration of the structural properties of starch components by the lack of an isoform of starch branching enzyme in rice seeds,  J. Biol. Chem. ,  1993, vol.  268 (pg.  19084- 19091) Google Scholar PubMed  Mizuno K,  Kobayashi E,  Tachibana M,  Kawasaki T,  Fujimura T,  Funane K, et al.  Characterization of an isoform of rice starch branching enzyme, RBE4, in developing seeds,  Plant Cell Physiol. ,  2001, vol.  42 (pg.  349- 357) Google Scholar CrossRef Search ADS PubMed  Nakamura Y.  Towards a better understanding of the metabolic system for amylopectin biosynthesis in plants: rice endosperm as a model tissue,  Plant Cell Physiol. ,  2002, vol.  43 (pg.  718- 725) Google Scholar CrossRef Search ADS PubMed  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 mutant of rice,  Plant J. ,  1997, vol.  12 (pg.  143- 153) Google Scholar CrossRef Search ADS   Nakamura Y,  Takeichi T,  Kawagushi K,  Yamanouchi H.  Purification of two forms of starch branching enzyme (Q-enzyme) from developing rice endosperm,  Physiol. Plant. ,  1992, vol.  84 (pg.  329- 335) Google Scholar CrossRef Search ADS   Nakamura Y,  Umemoto T,  Takahata Y,  Komae K,  Amano E,  Satoh H.  Changes in structure of starch and enzyme activities affected by sugary mutations in developing rice endosperm. Possible role of starch debranching enzyme (R-enzyme) in amylopectin biosynthesis,  Physiol. Plant. ,  1996, vol.  97 (pg.  491- 498) Google Scholar CrossRef Search ADS   Nakamura Y,  Yamanouchi H.  Nucleotide sequence of a cDNA encoding starch-branching enzyme, or Q-enzyme I, from rice endosperm,  Plant Physiol. ,  1992, vol.  99 (pg.  1265- 1266) Google Scholar CrossRef Search ADS PubMed  Nishi A,  Nakamura Y,  Tanaka N,  Satoh H.  Biochemical and genetic analysis of the effects of Amylose-Extender mutation in rice endosperm,  Plant Physiol. ,  2001, vol.  127 (pg.  459- 472) Google Scholar CrossRef Search ADS PubMed  O’Shea MG,  Morell MK.  High resolution slab gel electrophoresis of 8-amino-1,3, 6-pyrenetrisulfonic acid (APTS) tagged oligosaccharides using a DNA sequencer,  Electrophoresis ,  1996, vol.  17 (pg.  681- 686) Google Scholar CrossRef Search ADS PubMed  Ohdan T,  Francisco PBJr,  Sawada T,  Hirose T,  Terao T,  Satoh H, et al.  Expression profiling of genes involved in starch synthesis in sink and source organs of rice,  J. Exp. Bot. ,  2005, vol.  56 (pg.  3229- 3244) Google Scholar CrossRef Search ADS PubMed  Satoh H,  Nishi A,  Fujita N,  Kubo A,  Nakamura Y,  Kawasaki T, et al.  Isolation and characterization of starch mutants in rice,  J. Appl. Glycosci. ,  2003, vol.  50 (pg.  225- 230) Google Scholar CrossRef Search ADS   Satoh H,  Nishi A,  Yamashita K,  Takemoto Y,  Tanaka Y,  Hosaka Y, et al.  Starch-branching enzyme I-deficient mutation specifically affects the structure and properties of starch in rice endosperm,  Plant Physiol. ,  2003, vol.  133 (pg.  1111- 1121) Google Scholar CrossRef Search ADS PubMed  Satoh H,  Omura T.  New endosperm mutations induced by chemical mutagenesis in rice, Oryza sativa L,  Jpn. J. Breed. ,  1981, vol.  31 (pg.  316- 326) Google Scholar CrossRef Search ADS   Satoh H,  Shibahara K,  Tokunaga T,  Nishi A,  Tasaki M,  Hwang SK, et al.  Mutation of the plastidial α-glucan phosphorylase gene in rice affects the synthesis and structure of starch in the endosperm,  Plant Cell ,  2008, vol.  20 (pg.  1833- 1849) Google Scholar CrossRef Search ADS PubMed  Shapter FM,  Henry RJ,  Lee LS.  Endosperm and starch granule morphology in wild cereal relatives,  Plant Genet. Resour. ,  2008, vol.  6 (pg.  85- 97) Google Scholar CrossRef Search ADS   Svensson B.  Protein engineering in the α-amylase family: catalytic mechanism, substrate specificity, and stability,  Plant Mol. Biol. ,  1994, vol.  25 (pg.  141- 157) Google Scholar CrossRef Search ADS PubMed  Tanaka N,  Fujita N,  Nishi A,  Satoh H,  Hosaka Y,  Ugaki M, et al.  The structure of starch can be manipulated by changing the expression levels of starch branching enzyme IIb in rice endosperm,  Plant Biotechnol. J. ,  2004, vol.  2 (pg.  507- 516) Google Scholar CrossRef Search ADS PubMed  Tateoka T.  On the systematic significance of starch grains of seeds in Poaceae,  J. Jap. Bot. ,  1954, vol.  29 (pg.  341- 347) Tateoka T.  Further studies on starch grains of seeds in Poaceae from the view point of systematics,  J. Jap. Bot. ,  1955, vol.  30 (pg.  199- 208) Tateoka T.  Starch grains of endosperm in grass systematics,  Bot. Mag. Tokyo ,  1962, vol.  75 (pg.  377- 383) Google Scholar CrossRef Search ADS   Temnykh S,  Park WD,  Ayres N,  Cartinhour S,  Hauck N,  Lipovich L, et al.  Mapping and genome organization of microsatellite sequences in rice (Oryza sativa L.),  Theor. Appl. Genet. ,  2000, vol.  100 (pg.  697- 712) Google Scholar CrossRef Search ADS   Walker JT,  Merritt NR.  Genetic control of abnormal starch granules and high amylose content in a mutant of Glacier barley,  Nature ,  1969, vol.  221 (pg.  482- 483) Google Scholar CrossRef Search ADS   Yano M,  Okuno K,  Kawakami J,  Satoh H,  Omura T.  High amylose mutants of rice, Oryza sativa L,  Theor. Appl. Genet. ,  1985, vol.  69 (pg.  253- 257) Google Scholar CrossRef Search ADS PubMed  Yun MS,  Kawagoe Y.  Amyloplast division progresses simultaneously at multiple sites in the endosperm of rice,  Plant Cell Physiol. ,  2009, vol.  50 (pg.  1617- 1626) Google Scholar CrossRef Search ADS PubMed  © The Author 2010. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org

Journal

Plant and Cell Physiology – Oxford University Press

Published: Apr 1, 2010

Keywords: Amylopectin Endosperm Oryza sativa Rice Starch grain Thin section

There are no references for this article.